Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent
-Here we present Drucker's AdaBoost tailored for regression.
+Gradient boosting is again a similar technique to Adaptive boosting,
+it combines so-called weak classifiers or regressors into a strong
+method via a series of iterations.
-In bagging, each training example is equally likely to be
-picked. In boosting, the probability of a particular
-example being in the training set of a particular machine
-depends on the performance of the prior machines on
-that example. The following is a modification of
-Adaboost by Drucker.
-
-
-Start by selecting a set of training data \( n \) and assign to each entry a weight \( w_i=1 \) for \( i=1,2,\dots,n \). As we have done earlier, we could pick say \( 80\% \) of the data set for training. The algorithm runs as follows:
-
-
-
We define the probability that the training sample \( i \) is in the set by \( p_i = w_i/\sum_iw_i \). We pick \( n \) samples (with replacement) to form our training set. We pick a number uniformly in the range \( [0,\sum_iw_i] \).
-
We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.
-
Using every member of the training set with the chosen regression machine we obtain then a prediction \( \tilde{y}_i \).
-
We calculate then the loss function \( L_i \) for each training sample. We can use various types of loss function as long as we have a value
-
-
-\( L_i\in [0,1] \).
+In order to understand the method, let us illustrate its basics by
+bringing back the essential steps in linear regression, where our cost
+function was the least squares function.
@@ -239,7 +225,6 @@ Start by selecting a set of training data \( n \) and assign to each entry a wei
-Gradient boosting is again a similar technique to Adaptive boosting,
-it combines so-called weak classifiers or regressors into a strong
-method via a series of iterations.
+We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
+This means that for every iteration, we need to optimize
+
+$$
+(\hat{\boldsymbol{f}}) = \mathrm{argmin}_{\boldsymbol{f}}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
-In order to understand the method, let us illustrate its basics by
-bringing back the essential steps in linear regression, where our cost
-function was the least squares function.
+We define a real function \( h_m(x) \) that defines our final function \( f_M(x) \) as
+$$
+f_M(x) = \sum_{m=0}^M h_m(x).
+$$
+
+
+In the steepest decent approach we approximate \( h_m(x) = -\rho_m g_m(x) \), where \( \rho_m \) is a scalar and \( g_m(x) \) the gradient defined as
+$$
+g_m(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{m-1}(x_i)}.
+$$
+
+
+With the new gradient we can update \( f_m(x) = f_{m-1}(x) -\rho_m g_m(x) \). Using the above squared-error function we see that
+the gradient is \( g_m(x_i) = -2(y_i-f(x_i)) \).
+
+
+Choosing \( f_0(x)=0 \) we obtain \( g_m(x) = -2y_i \) and inserting this into the minimization problem for the cost function we have
+$$
+(\rho_1) = \mathrm{argmin}_{\rho}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i+2\rho y_i)^2.
+$$
@@ -225,7 +244,6 @@ function was the least squares function.
-We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
-This means that for every iteration, we need to optimize
-
+Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
$$
-(\hat{\boldsymbol{f}}) = \mathrm{argmin}_{\boldsymbol{f}}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+f_1(x) = f_{0}(x) -\rho_1 g_1(x)=-y_i.
$$
-
-We define a real function \( h_m(x) \) that defines our final function \( f_M(x) \) as
+We can then proceed and compute
$$
-f_M(x) = \sum_{m=0}^M h_m(x).
+g_2(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i,
$$
-
-In the steepest decent approach we approximate \( h_m(x) = -\rho_m g_m(x) \), where \( \rho_m \) is a scalar and \( g_m(x) \) the gradient defined as
-$$
-g_m(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{m-1}(x_i)}.
-$$
-
-
-With the new gradient we can update \( f_m(x) = f_{m-1}(x) -\rho_m g_m(x) \). Using the above squared-error function we see that
-the gradient is \( g_m(x_i) = -2(y_i-f(x_i)) \).
-
-
-Choosing \( f_0(x)=0 \) we obtain \( g_m(x) = -2y_i \) and inserting this into the minimization problem for the cost function we have
-$$
-(\rho_1) = \mathrm{argmin}_{\rho}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i+2\rho y_i)^2.
-$$
+and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \( m=M \). We can modify the steepest descent method, or steepest boosting, by introducing what is called gradient boosting.
-Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
-$$
-f_1(x) = f_{0}(x) -\rho_1 g_1(x)=-y_i.
-$$
-
-We can then proceed and compute
-$$
-g_2(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i,
-$$
-
-and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \( m=M \). We can modify the steepest descent method, or steepest boosting, by introducing what is called gradient boosting.
+Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
+so we do not learn a function that can generalize. However, we can modify the algorithm by
+fitting a weak learner to approximate the negative gradient signal.
+Suppose we have a cost function \( C(f)=\sum_{i=0}^{n-1}L(y_i, f(x_i)) \) where \( y_i \) is our target and \( f(x_i) \) the function which is meant to model \( y_i \). The above cost function could be our standard squared-error function
+$$
+C(\boldsymbol{y},\boldsymbol{f})=\sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
+
+
+The way we proceed in an iterative fashion is to
+
+
+
Initialize our estimate \( f_0(x) \).
+
For \( m=1:M \), we
+
+
+
compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
+
fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
+
update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
+
+
+
The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
+
+
@@ -226,7 +238,6 @@ and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \(
-Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
-so we do not learn a function that can generalize. However, we can modify the algorithm by
-fitting a weak learner to approximate the negative gradient signal.
+
+
-Suppose we have a cost function \( C(f)=\sum_{i=0}^{n-1}L(y_i, f(x_i)) \) where \( y_i \) is our target and \( f(x_i) \) the function which is meant to model \( y_i \). The above cost function could be our standard squared-error function
-$$
-C(\boldsymbol{y},\boldsymbol{f})=\sum_{i=0}^{n-1}(y_i-f(x_i))^2.
-$$
-
-
-The way we proceed in an iterative fashion is to
-
-
-
Initialize our estimate \( f_0(x) \).
-
For \( m=1:M \), we
-
-
-
compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
-
fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
-
update the estimate \( f_m(x) = f_{m-1}(x)+\nu h_m(u_m,x) \);
-
-
-
The final estimate is then \( f_M(x) = \sum_{m=1}^M\nu h_m(u_m,x) \).
-
-
@@ -238,7 +260,6 @@ The way we proceed in an iterative fashion is to
+XGBoost or Extreme Gradient
+Boosting, is an optimized distributed gradient boosting library
+designed to be highly efficient, flexible and portable. It implements
+machine learning algorithms under the Gradient Boosting
+framework. XGBoost provides a parallel tree boosting that solve many
+data science problems in a fast and accurate way. See the article by Chen and Guestrin.
-
-
+The authors design and build a highly scalable end-to-end tree
+boosting system. It has a theoretically justified weighted quantile
+sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning.
-# Load the data
-cancer = load_breast_cancer()
+
+It is now the algorithm which wins essentially all ML competitions!!!
-X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
-print(X_train.shape)
-print(X_test.shape)
-#now scale the data
-fromsklearn.preprocessingimport StandardScaler
-scaler = StandardScaler()
-scaler.fit(X_train)
-X_train_scaled = scaler.transform(X_train)
-X_test_scaled = scaler.transform(X_test)
-
-gd_clf = GradientBoostingClassifier(max_depth=3, n_estimators=100, learning_rate=1.0)
-gd_clf.fit(X_train_scaled, y_train)
-#Cross validation
-accuracy = cross_validate(gd_clf,X_test_scaled,y_test,cv=10)['test_score']
-print(accuracy)
-print("Test set accuracy with Random Forests and scaled data: {:.2f}".format(gd_clf.score(X_test_scaled,y_test)))
-
-importscikitplotasskplt
-y_pred = gd_clf.predict(X_test_scaled)
-skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
-save_fig("gdclassiffierconfusion")
-plt.show()
-y_probas = gd_clf.predict_proba(X_test_scaled)
-skplt.metrics.plot_roc(y_test, y_probas)
-save_fig("gdclassiffierroc")
-plt.show()
-skplt.metrics.plot_cumulative_gain(y_test, y_probas)
-save_fig("gdclassiffiercgain")
-plt.show()
-
-XGBoost or Extreme Gradient
-Boosting, is an optimized distributed gradient boosting library
-designed to be highly efficient, flexible and portable. It implements
-machine learning algorithms under the Gradient Boosting
-framework. XGBoost provides a parallel tree boosting that solve many
-data science problems in a fast and accurate way. See the article by Chen and Guestrin.
-
-The authors design and build a highly scalable end-to-end tree
-boosting system. It has a theoretically justified weighted quantile
-sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning.
+
+
+As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.
-In bagging, each training example is equally likely to be
-picked. In boosting, the probability of a particular
-example being in the training set of a particular machine
-depends on the performance of the prior machines on
-that example. The following is a modification of
-Adaboost by Drucker.
-
-
-Start by selecting a set of training data \( n \) and assign to each entry a weight \( w_i=1 \) for \( i=1,2,\dots,n \). As we have done earlier, we could pick say \( 80\% \) of the data set for training. The algorithm runs as follows:
-
-
-
We define the probability that the training sample \( i \) is in the set by \( p_i = w_i/\sum_iw_i \). We pick \( n \) samples (with replacement) to form our training set. We pick a number uniformly in the range \( [0,\sum_iw_i] \).
-
We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.
-
Using every member of the training set with the chosen regression machine we obtain then a prediction \( \tilde{y}_i \).
-
We calculate then the loss function \( L_i \) for each training sample. We can use various types of loss function as long as we have a value
-
-
-
-\( L_i\in [0,1] \).
-
-
-
-
-
Gradient boosting: Basics with Steepest Descent
+
Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent
Gradient boosting is again a similar technique to Adaptive boosting,
@@ -1051,7 +1022,7 @@ function was the least squares function.
-
The Squared-Error again! Steepest Descent
+
The Squared-Error again! Steepest Descent
We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
@@ -1094,7 +1065,7 @@ $$
-
Steepest Descent Example
+
Steepest Descent Example
Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
@@ -1116,7 +1087,7 @@ and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \(
-
Gradient Boosting, algorithm
+
Gradient Boosting, algorithm
Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
@@ -1141,15 +1112,15 @@ The way we proceed in an iterative fashion is to
compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
-
update the estimate \( f_m(x) = f_{m-1}(x)+\nu h_m(u_m,x) \);
+
update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
-
The final estimate is then \( f_M(x) = \sum_{m=1}^M\nu h_m(u_m,x) \).
+
The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
-
Gradient Boosting, Examples of Regression
+
Gradient Boosting, Examples of Regression
@@ -1204,7 +1175,7 @@ plt.show()
-
Gradient Boosting, Classification Example
+
Gradient Boosting, Classification Example
@@ -1253,7 +1224,7 @@ plt.show()
-
XGBoost: Extreme Gradient Boosting
+
XGBoost: Extreme Gradient Boosting
XGBoost or Extreme Gradient
@@ -1274,7 +1245,7 @@ It is now the algorithm which wins essentially all ML competitions!!!
-
Regression Case
+
Regression Case
@@ -1330,7 +1301,7 @@ plt.show()
-
Xgboost on the Cancer Data
+
Xgboost on the Cancer Data
As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.
diff --git a/doc/pub/week45/html/week45-solarized.html b/doc/pub/week45/html/week45-solarized.html
index d0c53209e..8bac88399 100644
--- a/doc/pub/week45/html/week45-solarized.html
+++ b/doc/pub/week45/html/week45-solarized.html
@@ -78,28 +78,28 @@ div { text-align: justify; text-justify: inter-word; }
'___sec19'),
('Basic Steps of AdaBoost', 2, None, '___sec20'),
('AdaBoost Examples', 2, None, '___sec21'),
- ('Additive boosting for Regression', 2, None, '___sec22'),
- ('Gradient boosting: Basics with Steepest Descent',
+ ('Gradient boosting: Basics with Steepest Descent/Functional '
+ 'Gradient Descent',
2,
None,
- '___sec23'),
+ '___sec22'),
('The Squared-Error again! Steepest Descent',
2,
None,
- '___sec24'),
- ('Steepest Descent Example', 2, None, '___sec25'),
- ('Gradient Boosting, algorithm', 2, None, '___sec26'),
+ '___sec23'),
+ ('Steepest Descent Example', 2, None, '___sec24'),
+ ('Gradient Boosting, algorithm', 2, None, '___sec25'),
('Gradient Boosting, Examples of Regression',
2,
None,
- '___sec27'),
+ '___sec26'),
('Gradient Boosting, Classification Example',
2,
None,
- '___sec28'),
- ('XGBoost: Extreme Gradient Boosting', 2, None, '___sec29'),
- ('Regression Case', 2, None, '___sec30'),
- ('Xgboost on the Cancer Data', 2, None, '___sec31')]}
+ '___sec27'),
+ ('XGBoost: Extreme Gradient Boosting', 2, None, '___sec28'),
+ ('Regression Case', 2, None, '___sec29'),
+ ('Xgboost on the Cancer Data', 2, None, '___sec30')]}
end of tocinfo -->
@@ -141,7 +141,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
-In bagging, each training example is equally likely to be
-picked. In boosting, the probability of a particular
-example being in the training set of a particular machine
-depends on the performance of the prior machines on
-that example. The following is a modification of
-Adaboost by Drucker.
-
-
-Start by selecting a set of training data \( n \) and assign to each entry a weight \( w_i=1 \) for \( i=1,2,\dots,n \). As we have done earlier, we could pick say \( 80\% \) of the data set for training. The algorithm runs as follows:
-
-
-
We define the probability that the training sample \( i \) is in the set by \( p_i = w_i/\sum_iw_i \). We pick \( n \) samples (with replacement) to form our training set. We pick a number uniformly in the range \( [0,\sum_iw_i] \).
-
We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.
-
Using every member of the training set with the chosen regression machine we obtain then a prediction \( \tilde{y}_i \).
-
We calculate then the loss function \( L_i \) for each training sample. We can use various types of loss function as long as we have a value
-
-
-\( L_i\in [0,1] \).
-
-
-
-
-
Gradient boosting: Basics with Steepest Descent
+
Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent
Gradient boosting is again a similar technique to Adaptive boosting,
@@ -975,7 +947,7 @@ function was the least squares function.
-
The Squared-Error again! Steepest Descent
+
The Squared-Error again! Steepest Descent
We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
@@ -1010,7 +982,7 @@ $$
-
Steepest Descent Example
+
Steepest Descent Example
Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
@@ -1028,7 +1000,7 @@ and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \(
-
Gradient Boosting, algorithm
+
Gradient Boosting, algorithm
Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
@@ -1051,15 +1023,15 @@ The way we proceed in an iterative fashion is to
compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
-
update the estimate \( f_m(x) = f_{m-1}(x)+\nu h_m(u_m,x) \);
+
update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
-
The final estimate is then \( f_M(x) = \sum_{m=1}^M\nu h_m(u_m,x) \).
+
The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
-
Gradient Boosting, Examples of Regression
+
Gradient Boosting, Examples of Regression
@@ -1113,7 +1085,7 @@ plt.show()
-
Gradient Boosting, Classification Example
+
Gradient Boosting, Classification Example
@@ -1161,7 +1133,7 @@ plt.show()
-
XGBoost: Extreme Gradient Boosting
+
XGBoost: Extreme Gradient Boosting
XGBoost or Extreme Gradient
@@ -1182,7 +1154,7 @@ It is now the algorithm which wins essentially all ML competitions!!!
-
Regression Case
+
Regression Case
@@ -1237,7 +1209,7 @@ plt.show()
-
Xgboost on the Cancer Data
+
Xgboost on the Cancer Data
As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.
diff --git a/doc/pub/week45/html/week45.html b/doc/pub/week45/html/week45.html
index 2419732b3..920dd8dce 100644
--- a/doc/pub/week45/html/week45.html
+++ b/doc/pub/week45/html/week45.html
@@ -83,28 +83,28 @@ div { text-align: justify; text-justify: inter-word; }
'___sec19'),
('Basic Steps of AdaBoost', 2, None, '___sec20'),
('AdaBoost Examples', 2, None, '___sec21'),
- ('Additive boosting for Regression', 2, None, '___sec22'),
- ('Gradient boosting: Basics with Steepest Descent',
+ ('Gradient boosting: Basics with Steepest Descent/Functional '
+ 'Gradient Descent',
2,
None,
- '___sec23'),
+ '___sec22'),
('The Squared-Error again! Steepest Descent',
2,
None,
- '___sec24'),
- ('Steepest Descent Example', 2, None, '___sec25'),
- ('Gradient Boosting, algorithm', 2, None, '___sec26'),
+ '___sec23'),
+ ('Steepest Descent Example', 2, None, '___sec24'),
+ ('Gradient Boosting, algorithm', 2, None, '___sec25'),
('Gradient Boosting, Examples of Regression',
2,
None,
- '___sec27'),
+ '___sec26'),
('Gradient Boosting, Classification Example',
2,
None,
- '___sec28'),
- ('XGBoost: Extreme Gradient Boosting', 2, None, '___sec29'),
- ('Regression Case', 2, None, '___sec30'),
- ('Xgboost on the Cancer Data', 2, None, '___sec31')]}
+ '___sec27'),
+ ('XGBoost: Extreme Gradient Boosting', 2, None, '___sec28'),
+ ('Regression Case', 2, None, '___sec29'),
+ ('Xgboost on the Cancer Data', 2, None, '___sec30')]}
end of tocinfo -->
@@ -146,7 +146,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
-In bagging, each training example is equally likely to be
-picked. In boosting, the probability of a particular
-example being in the training set of a particular machine
-depends on the performance of the prior machines on
-that example. The following is a modification of
-Adaboost by Drucker.
-
-
-Start by selecting a set of training data \( n \) and assign to each entry a weight \( w_i=1 \) for \( i=1,2,\dots,n \). As we have done earlier, we could pick say \( 80\% \) of the data set for training. The algorithm runs as follows:
-
-
-
We define the probability that the training sample \( i \) is in the set by \( p_i = w_i/\sum_iw_i \). We pick \( n \) samples (with replacement) to form our training set. We pick a number uniformly in the range \( [0,\sum_iw_i] \).
-
We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.
-
Using every member of the training set with the chosen regression machine we obtain then a prediction \( \tilde{y}_i \).
-
We calculate then the loss function \( L_i \) for each training sample. We can use various types of loss function as long as we have a value
-
-
-\( L_i\in [0,1] \).
-
-
-
-
-
Gradient boosting: Basics with Steepest Descent
+
Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent
Gradient boosting is again a similar technique to Adaptive boosting,
@@ -980,7 +952,7 @@ function was the least squares function.
-
The Squared-Error again! Steepest Descent
+
The Squared-Error again! Steepest Descent
We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
@@ -1015,7 +987,7 @@ $$
-
Steepest Descent Example
+
Steepest Descent Example
Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
@@ -1033,7 +1005,7 @@ and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \(
-
Gradient Boosting, algorithm
+
Gradient Boosting, algorithm
Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
@@ -1056,15 +1028,15 @@ The way we proceed in an iterative fashion is to
compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
-
update the estimate \( f_m(x) = f_{m-1}(x)+\nu h_m(u_m,x) \);
+
update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
-
The final estimate is then \( f_M(x) = \sum_{m=1}^M\nu h_m(u_m,x) \).
+
The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
-
Gradient Boosting, Examples of Regression
+
Gradient Boosting, Examples of Regression
@@ -1118,7 +1090,7 @@ plt.show()
-
Gradient Boosting, Classification Example
+
Gradient Boosting, Classification Example
@@ -1166,7 +1138,7 @@ plt.show()
-
XGBoost: Extreme Gradient Boosting
+
XGBoost: Extreme Gradient Boosting
XGBoost or Extreme Gradient
@@ -1187,7 +1159,7 @@ It is now the algorithm which wins essentially all ML competitions!!!
-
Regression Case
+
Regression Case
@@ -1242,7 +1214,7 @@ plt.show()
-
Xgboost on the Cancer Data
+
Xgboost on the Cancer Data
As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.
diff --git a/doc/pub/week45/ipynb/ipynb-week45-src.tar.gz b/doc/pub/week45/ipynb/ipynb-week45-src.tar.gz
index 123bc0208..6c4948678 100644
Binary files a/doc/pub/week45/ipynb/ipynb-week45-src.tar.gz and b/doc/pub/week45/ipynb/ipynb-week45-src.tar.gz differ
diff --git a/doc/pub/week45/ipynb/week45.ipynb b/doc/pub/week45/ipynb/week45.ipynb
index 316a9c68a..60faab4e4 100644
--- a/doc/pub/week45/ipynb/week45.ipynb
+++ b/doc/pub/week45/ipynb/week45.ipynb
@@ -10,7 +10,7 @@
" \n",
"**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n",
"\n",
- "Date: **Nov 6, 2020**\n",
+ "Date: **Nov 12, 2020**\n",
"\n",
"Copyright 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n",
"\n",
@@ -77,7 +77,9 @@
{
"cell_type": "code",
"execution_count": 1,
- "metadata": {},
+ "metadata": {
+ "collapsed": false
+ },
"outputs": [],
"source": [
"%matplotlib inline\n",
@@ -131,21 +133,10 @@
{
"cell_type": "code",
"execution_count": 2,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhsAAADyCAYAAADk8wfeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydd3gVxfrHP3Nqeq8EEkLovYOCCNYrIBbs5Vq59p8VKygWlGu5dkW99oJiAxUUqVJEmghIqKGl936SnDa/PzY5JTknBSHAdT7Pkye7szOzs5uTs999533fEVJKFAqFQqFQKI4WumM9AIVCoVAoFP/bKLGhUCgUCoXiqKLEhkKhUCgUiqOKEhsKhUKhUCiOKkpsKBQKhUKhOKoosaFQKBQKheKoYjjWAzhe0IVHSn1Coms/tcpJgNO7jikppJ1HpVAoFArF0WPTpk1FUsrYo30eJTbq0SckEj37M9f+q2ur6VPhVhtBQ+OJuqj7sRiaQqFQKBRHBSHEwfY4T7tOowghooQQ3wohqoUQB4UQV/ipd60QwiGEqPL4GetxvKrRj0MI8Wr9sc5CCNno+PS2jnV+R6PXvjE2sK1dKBQKhUKhoP0tG68DViAeGAgsEEJskVJu91F3rZRytK9OpJSu+QwhRDCQD3zZqFqElNJ+uANdF20A6jzO6b+udEqyH15N2JkphJ2efLinVCgUCoXif5J2s2zUi4LJwHQpZZWUcjXwHXD1X+z6IqAAWPUX+/EiO6jRrWlObNi16ZaKZYeO5BAUCoVCofifoD2nUboDDinlbo+yLUAfP/UHCSGKhBC7hRDThRD+rDDXAB/Jpou8HBRCZAkh3hdCxPhqKIT4lxBioxBiY4ujdzajNurFBhKyZ6ylYkVmi90pFAqFQvF3oT3FRghQ3qisHAj1UXcl0BeIQ7OGXA5MbVxJCJEMnAp86FFcBAwDUoAh9f1/6mtAUsq3pZRDpZRDWxq8bEZsWHOrtQ2nRNbaqfjpQEvdKRQKhULxt6E9xUYVENaoLAyobFxRSrlPSrlfSumUUm4DnkCbLmnMP4HVUsr9Hm2rpJQbpZR2KWU+cDtwlhCi8bnbRjNiQ4kLhUKhUCj8055iYzdgEEJ08ygbAPhyDm2MBISP8n/ibdXw1xY/7VuNdPgXG9LmaFrW3LSLQqFQKBR/I9pNbEgpq4FvgCeEEMFCiFHAecDHjesKIc4RQsTXb/cEpgPzG9U5GUiiURSKEGKEEKKHEEInhIgGXgFWSCkbT+G0iaqVWX6P2fIsTcqcFttfOZ1CoVAoFP8ztHe68luBQLTokTnALVLK7UKI5Pp8GA1xo6cDW4UQ1cBCNJHydKO+rgG+kVI2nobpAvyENj3zJ1r86uVH5WqaQdqcLVdSKBQKheJvQLvm2ZBSlgDn+yg/hOZA2rB/H3BfC33d5Kd8DpqQOWrY8qoxRAcgjHq/dapWZRMxKe1oDkOhUCgUihMCtRCbH87K1aZBTJ28g2WcFhv5L/1OyVd7mm3fNBJXoVAoFIq/J0ps+KFbpTYNEntzfwB0wUZyn1lHbYbm+mE90LwLSPXa3KM7QIVCoVAoThCU2PCDsz52Reh1GJNCcFbbcJRbqVhSv2aN+EvBLQqFQqFQ/G1QYsMPXpMgHmGs9vz6yJN6raGmSxQKhUKhaB4lNvzg9DBc2BoyhHrSYNmwK7GhUCgUCkVzKLHhh5YkRIPWaFiErYEGHw+FQqFQKBQa7b3E/AmDswWXDHtxLVkPrnKFt4adkYw+woy5c3g7jE6hUCgUihMHZdnwYJL82rXd2pRcZd9lAKAPNxM8NMHrWF0LESsKhUKhUPwdUGLDAwMeKcbbGGxi+aPAtW1M0vKTlf2w70gMS6FQKBSKExolNvzQ5mTjOrc60UeYAbBlVSEdLfdky6+mam2Oa99RZaVg9hYc5XVtHYVCoVAoFMcdSmx44GnMcApB2FkprW4bempHdz8G920tbSHTKEDBa39QNj8D6ZBYs6vIfWod1gMVVP2a02JbhUKhUCiOd5TY8CABd9ZPCRjjglrd1pTkTmsu9G7ZYtlc4Ku6+/jWQteibWXz9lL8cbrrWOUv/leaVSgUCoXiREGJDQ/CKGfa9pcIElKLRmlDllBhcNc1p0U0W9dZa6d6fR5ZD66i5LOdrvLqDXk4yrynTlTSMIVCoVCc6Cix0RihPeCloG1Oonr3rQwaHOe3mqPaRs6MtZR+0/L0CkDlssw2DEKhUCgUiuMPJTZ8UIOOOSkmTXAAHR4d2WIb4eEgKhpZRKTDbZ0o+mB7i33pI8wYogMAqFh8EGlztGbYCoVCoVAclyix0Qy7HFoorC7I2Oa2ntYNT7Fgy6z0WT+wf4xr2xATiL241rXvqLa3+fwKhUKhUBwvqAyizVDl4S8Re1N/bHnVlM3PaFXbyIu6g1Ni+UNzAC1dtNflCNqYwH4xRF/Ri2J2ULO1CF2wt7iRtXbAfNjXoVAoFArFsURZNprB4TEbYk4NJ+SkDj7rhZyS1KRM6ITLUdSaVUn12lwsG/N9to84V0t5rjNr2k8YdERc0NU9jmqbz3YKhUKhUJwIKMtGM9h8LMcWf/dgpF1S8OpmTMmhxFzbBxHo+zYKo6blij9M93m8AX2YCYCancUAWDblE3Vxd0ydQil4ZTNOi5pGUSgUCsWJixIbzfBSeTln4m3NMMYHA5D48AhEgB6dSe+3vbS2nD1UmN3tXfXrLSoNviJl3+0lqF9M46YKhUKhUJwQtOs0ihAiSgjxrRCiWghxUAhxhZ961wohHEKIKo+fsR7HVwghaj2O7WrU/nQhxE4hhEUIsVwI0apUoAKJ8LBmbLP6TxeuDzM1KzQAnNaWo0gSpg51bUdd1gOA+DsHa+eo991wVqppFIVCoVCcuLS3z8brgBWIB64E3hRC9PFTd62UMsTjZ0Wj47d7HOvRUCiEiAG+AaYDUcBG4IvDGexfDTj1TFvui/BzUtGHmFz7gb2i6TjrFIwJmvWkYRoGwJpbrVaRVSgUCsUJSbtNowghgoHJQF8pZRWwWgjxHXA18OARPNWFwHYp5Zf1550BFAkhekopdzbbshFtXoytEbog37dXH2Yi/u4h6Pz4evii4OXfAUh4aDiGcBWZolAoFIoTh/a0bHQHHFLK3R5lWwB/lo1BQogiIcRuIcR0IUTjJ/Mz9cfXeE6x1Pe3pWFHSlkNZPg6jxDiX0KIjUKIjb4G8FcThQf68bOIvKRHm4SGJzXbilSSL4VCoVCcULSn2AgBGs8DlAOhPuquBPoCcWjWkMuBqR7HHwC6AEnA28D3Qoi0tp5HSvm2lHKolHJo42MNlJWV+TvUIp6ZRCMv7g6ALsRIQNfm107xJO6OQV775T/so2D21sMek0KhUCgU7U17io0qIKxRWRjQJKWmlHKflHK/lNIppdwGPAFc5HF8nZSyUkpZJ6X8EFgDjG/reVqiX9ZeMjJal8SrJQK6RtBx1il0mNZy6nNPTEkhJE4fScS5XVxltuwqQFvDxbKtEOn4qxM+CoVCoVAcPdpTbOwGDEKIbh5lA4CWFwvRZjSaWxbN8/j2+n4Bl69IWivP40Wkpc36xC/6v+BnoQ82EjLKO3GYZWshtbtKKfl0J9mPrPmrw1MoFAqF4qjRbmKj3nfiG+AJIUSwEGIUcB7wceO6QohzhBDx9ds90SJL5tfvRwghzhZCBAghDEKIK4ExwKL65t8CfYUQk4UQAcCjwNa2OofWj6TtTY4ixsRg13bJZztxlNU2U1uhUCgUiuOD9g59vRUIBAqAOcAtUsrtQojk+nwZyfX1Tge2CiGqgYVoIuXp+mNG4CmgECgC7gDOl1LuApBSFqL5ecwESoERwGWHM1inaCo29hVW8eLi3UjZOvfRxGkjSHx4xOGcvgnxdw4m8iJPw5B7fHYlPBQKhUJxnNKuGUSllCXA+T7KD6E5djbs3wfc56ePQmBYC+dZAvT8S4P1w9Xvrie7rIarRqYQG9ry1IhnHo0jQfDQBEq/2gNA2by9rvK8WRvoOOuUI3ouhUKhUCiOBGohtmaQPiwbtcdB2GnI6KYLvwHUpBe380gUCoVCoWgZJTaaQSLY90dho7KG3381C8fh46z1vTBb5YrMdh6JQqFQKBQto8RGM0ghOPhnsZd/Rkm1FQCH89iJDc+l6iMv6U7iI5pPiLkN+TsUCoVCoWgvlNjwgcGhWQ4kUBuYh7NeWOzMq3DVsdmPndiIv3uwa9vcORx9qAlh1FGzreiYjUmhUCgUCn8oseGDS7YsBTTLht1Uya6d2qKyDVYNAOsxTKTVsMw9aBlJAaTNib2wxu8Ui0KhUCgUxwolNnwQatXCSBscRNet+61JHdsxztqZ+MgIoq/u1WSZ+5wZa4/RiBQKhUKh8I0SGz7Q1U+bNEyUHDx0EICNB0pddeyOYzeNAqAPNRHYx73QW/R1/tazUygUCoXi2KLEhg909TKjcejrfxa7F6w9ltMovgjoGunatpeqBF8KhUKhOH5QYsMHg/Zpv2Uz6cqP9TRKY4ReYIgPAiDv3xso+mA7+a//gXRKpO34GqtCoVAo/l4oseGDZGskQkqfSb0aaCw23l+znxs/3HBMQ2LjbnGtP0ftzhJsmZVkP7ya7Olr1MqwCoVCoThmKLHhg7CkgZrYaFTev2O4a7u8xuYKiQV4/Pt0luwoIO3hhe00yqboAvxnn3dU2dpxJAqFQqFQuFFiwy8SGlk2hqS4/SJu/2wzLy/d096DOmzs+ZZjPQSFQqFQ/E1RYsMHEhCy6aqvjSNQ5v+R3Y6jah3x9wzBlBJGTKPolKL3/jxGI1IoFArF3x0lNnwiET7WPvn4t4Ne+2U1x9/UhDEuiLhbBhDQI4qEh4Z7HXOU1x2jUSkUCoXi70ybxYYQIkQIEdxyzRMXzbLh7SD6wqUTObnEO2FWmeX4ExueGMLNXsvO5z6znoqlh47hiBQKhULxd6TVYkMIcZsQ4hBQDlQIIQ4KIW49ekM7tgjAWR/6qq+rAWBI+R8+66bnVPgsP16ImdLPtV2x+OAJtxT9te+v56c/87A7nFz133Wc9sIKsstqjvWwFAqFQtFK/IcveCCEeBh4CHgeWF1ffAowSwgRJqWcdZTGd8wQ0u0gKuzNrzcy/pVVXvtSSkQzYbPtjaljqNd+8UfpdJx1Cs/+tJOhnSM5rWd8i31YrHakhGBzqz4yR4wXft7Fil2FrNhV6FU+atYyMp4ej153/NxnhUKhUPimtU+Om4F/SSnneJQtFULsAZ4G/qfEhkR6h74627a4mdXhxGzQt1yxndCZ9QSPSKB2bxmOYi276E+bs3ljRQYAK6eOIzk6yGfbuRszsdqdfPjrAWwOJyumjmu3cRdU1vLqsr1+j6c9vJDN088kMtgEwKtL9/BCfZbX5y8eQFJEICelRbvql1tsbM4sZWyPOL99llts3PLpJn7N0Kw/+54ejxAwZ30mlw7rpMSNQqFQHAatFRtxwAYf5euBll+LTySEJjEEbp8Nz0yiFw3pyFebsgAwGXReuTYa6DHtJ4anRjH3ppPaYcCtI/KCbgBkPahZYTK/2Ok6Nua55ax58DSSIgK92nR+cEGTfrZmldErMQyj/uj5FlfW2njrl328tryp0Hh8Uh8ig03835zNAAx6cjEvXjqALZnlfPDrAVe9+77cAsCLlw7ggkEdeebHHbz1i5Ya9uS0aD65YQS6RsJhzd4irvzvOq+yPQVVnP3SSgDeWpnBwWILE/on8voVg4/Y9SoUCsX/Oq19YuwGrvBRfgWw68gN5/igujgDIT1EhseUiFHv3rbanVT6WdJ9/f6SozrGwyXq0h4AnIPJq3zUrGV8uTHTtf/8Iu3PqpfgGZgz6bU1dHvkx6M6xse/T/cSGntmnsOBWRNIf+Js/nlSCpMGdOD+f/RwHb/7iy1eQsOTu7/YwqLteS6hAfBrRjFPL9zh2rc5nDz1Q7qX0Aipny5qEBoAB4u1XCULtuZSWm39axepUCgUfyNaKzZmAI8KIZYIIR4XQswQQiwBpgGPtfZkQogoIcS3QojqegdTXwIGIcS1QgiHEKLK42ds/TGzEOLd+vaVQojNQohzPNp2FkLIRm2nt3aMGhK9BFmvKxwh7syhBp33LdueW+7a7pcUTms4UFSNwynZklnG4CcXt6uDaW64EYB1zqaRNFO/2sqm1Vm8fvMyAr/NZmpZIPeUB3JptQnjX8jC/v6a/XR+cAG3ffa73zpOp6TX9J+45K21LssRQGpMsMuKEmQyuHxhbh3b1Wc/v0wdy6r7x/HL1LGusps+3gTAOX0TXGX/Xa2N6R8vraTbIz/y39X7XcfOHdCBbTPOavaaBj25mN8PlSLlsV39V6FQKE4EWiU2pJTfACOAPGAiMKl+e7iUcl4bzvc6YEWberkSeFMI4W9t9LVSyhCPnxX15QYgEzgVCAemA3OFEJ0btY/waPtkG8YICM2m4WhqtTDovU3vdXb3miMXDk5qsedDxRbGPr+Cs19ayXmvr6Gk2trEwfRo8tOzmym2OxmhM/Jh5w58N2kQSAh3CHQSfvtkd5M2yXY9d5V7T7H8fqi01ed8/Pt0QLMIPP79dgBu/ngTzy1yT+XsL66mxuZoYhF6+bKBfvs9MGsCH1w3jJcvG8gvU8fy2Y0jSIkOplNUECnRway639u/5M2rhrDv6fGYDO6P/c68Std2p6hA9s48h1cvH4QQgpcu1c49NCWSt68eAuDV54Vv/Mopzy5v9X1QKBSKvyutDi2QUm4CrjrcE9Xn5pgM9JVSVgGrhRDfAVcDD7ZhHNVolpYGfhBC7AeGAAcOd3yeVBkc3tMoHjT2VXjNw4HR1yJsC7bmcttnv7PknlPpGhdCXoXmoLm3oOpIDLVNVBRp4aLmeutA2oEq5v9RylQCm2vmYvFdY5j54w5W7Crk098OMTg5kh25FfRMCKWi1s7V764j0Kjn0XN706eDZuW58cONXn28v+YAFw7qyE/b82A7jOoaw8lpMWw84C0ytjx6FuFBxhbH5OnsmRLtnf6lU1QQK6eOY8xzy9k07QwAdDrBzif+wZSPNrJ0Z4Gr7kuXDuT8Qd5i8fxBSV5lB2ZNAODzf43ksrd/AyCrtIbODy7gsmGdeObCfsdVFJJCoVAcL/gVG0KIKCllScN2c5001GuB7oBDSun56rwFzULhi0FCiCKgBPgYeEZK2cTUIISIr+97e6NDB4UQElgMTJVSFvlo+y/gXwCG7r1c5T/H56Hzs+qrsZFlY9NB7Q3/8Ul92FfoLSB25Fa4pg5mfLedT24cgcXq28fjaIfLvrxkD4avNJ+MEI9rCBBQ62cmYPL9Q4iIC+Ld+zTLS/rne3l9ykD6z1xCQrjZJaQaM+GV1bx7zVBOSotmyY58AD64bhjXvq/5GJ/72mpX3SveWcf+Z8bzwNfbABjVNZpRXWNaJTRaQ3J0kEskNKDTCd69dhjfbs4iKSKIISmRbYoyGdklmgOzJvDMwh28tVLzBfl8QyYOp2TigA6c2j32iIxdoVAo/ldobhqlUAjR8NpYBBT6+Gkobw0haAnBPCkHQn3UXQn0RYuCmQxcDkxtXEkIYQQ+BT6UUjbY5IuAYUAKmrUjtL5OE6SUb0sph0ophzbpG99iw6DTsfPJf7j2G/w0ooJNdGgUzbHxoHuqIbushhqrg+nzfa9RYrE6fJYfCaSUvLjErfFWVLr9NTqZmn4Enouo4caXxpDQJZyAECNXzBgBQNbOUj64dzXhNnh9eUazPhg3fLiR3o8ucu2f2j2W28al+ayb+pC2Uu6IWgPj0q3ccqrvekeaCwZ1ZHhq1GGHs951Rnd6JYYRZNLCnL/clMU1763npGeWsm5fMS/8vIvXfUTU/C8ipST/1c3kPPUbtkILzlo79tLa9h2DUyLtTiqWHqLkq904W5vht7YCHMd3NmCF4kSnuWmU09CsCg3bf9UTrgoIa1QWBlQ2riil3Oexu00I8QSa2HimoVAIoUOzeFiB2z3aVgENtvt8IcTtQG598rFWe2I2nkaRQoeQTox6QYDRnUNjW7amn3RCcP3oVEICDDzyrSYops9zC4v9RdX0evQnv+e76eNNfHLjiNYOr9VU19np89gir7KuV3Zn2Se7OS3MSO9APXvqnHToFkHOnjKeC6+favFYrj4ywXt64obKAJ6LaJrBs3diGOm5TW/xl+MHUF5Qw9SzezK+XyITXlnNjaNTOatPApe8VZ8CXsKYWiN12HnjluXcNvu0v3rprUdK2LcCUseArvX5UQJNen68U0sH32/GIldkUm55LZfWT7MATOyfyPacCuxOyaQBHVrsd+OPB9j/RyEX3j+EkuxqDmwrYtiE1LZdUztiL68j75n1rv38Fza5thPuH4YhKuAv9e+sv6+6+s+klBJnhZXcZ9ZjiA4g/p6hZD+yukk7y8Z817YxKYSQkYmUfu1eqTnutoGYdr8Iq17QCm7bAL/8G06fDpGdj7vkfArFiYxfsSGl/MVje8URONduwCCE6CalbPiPH0DT6Q+fwwH3k19o3wDvojmajpdSNvda0iCSWvWtIVy/na5oFABbZCymknwM9T4b398+2ms6oMRixajXceWIFASCh7/d1prTcd9Z3Xn+592s3ttklqfNSCnZuPAAvUd1wBxqZNvaXK78Qcs30cfqfoiOPLkjI0/u6Mq5cfUd/QnrEwPARRW1BPhISDbqoq6s+cr9lh7rEBTqtVubGB7A+R2iCVlbwv0X92H7l5pWXBFgI8fgZP1nu1kPJPeOouvQeKaWBZK0p47hE92zc+NqvKdNcvaU0qFbZLPXm7OnDJ1BkJDauiggL5xOqC2DoCj45VlY8TQMvgYmvQJAbZWNnD1ldBnkMSVy8FcIiYfoppaXrY+dxeL0fP718aYmx059boVruyE/yFUjk7lyRAo9E0K9Hmhbl2eybr52/2bf5m63ffEOrgm/nDfyvwXg7Cl9WfTOn5x/9yCSejS9T1JK6qrtBITU31eHHUr2QV2l9kC94gt3SHfZITAGQ3B0k34aKPzvNur2lvk97o+8Z7Wps6SnRyM8LUhOJ1QXwDunQ0UWDLwK0ufDnX8gA6MpmbOTmm3e/xNhZ6ZQsdh7MUR7ca1PodEYW3aVl9AAKHrnNxLFC+7I9teHAWDdupkC68sABA2Ow7K1EOySgD7RRF/Wg4qlhxBGPWGnJ7flVigUf2tEa0L3hBAOIFFKWdCoPBookFK26nVQCPE52sP/RmAgsBA4WUq5vVG9c4DfpZT5QoiewFfAl1LKx+uPz65vf0a9JcOz7QigDNgDRAJvAHFSymZTXxp79JYvvNkNw85wCgq68PWQMwirLObM3drDwVSQhbk4j/DbX+HGU7pQZrEy8InFrvYzL+jLlSNSAKizO+gxzb8Vo4F/T+7HtuxyPvlNWxytsW9BW8naVcr8F7XxHhoRTvI6zeryXHgNU+ujSSITgrhixkit/oPuKJi4OwZhSgpptv/Niw/x69ea4Og2NI6Bl3YlLlR7a3395mVtHm//cR3ZujyL2lA9AZW+p5E8x+tJVWktHz70KwC3zT4NZ40dYdAhjD5mBi0l4LBCaALMvQb6XwKf+4y6hh4TKMsp5dM92qzduKt60mNkAuW7dxI1tz5J29R9fh/MNVaH6+FlNuhcU0QtEeoUdLLrmGAxtVzZD8PPTUUIQf7+cg5sc69/MzH1Q3Rl++lk3uKuHNYRRt1Zf0+u1soarit7E+RugQ6DICSBmq9ep3j32c2eWxdkIHHaSOocTix1doY8tYRnCeQk3CLSEVJBZMRGggo+xaDTrA5ltuuockwGIEQ/jzDDHHLqvjis608yn48Qbp8oKXWU2v4PvSig0nGFV71y051UVWpfCcIskHUNyfwsSHxn0/VF2BnJhJ2RcljjVSiOB4QQm3y5EhxpWhuN4s8qYEabxmgttwLvAQVAMXCLlHK7ECIZSAd6SykPAacDHwghQoB84BO0tOgIIVKAm4A6IM/jrfAmKeWnQJf6unFABZqD6OWtH6JnBlHPcm2nIRrF0CgqRXjcopZSlU85JZVHJvQGYGSXapfYWLA1lwn9E1s/1EZYa9xftFu2FZJcn7jrlgq3GfvyR91TNRHnp1E2T0tZXvDqZq8VYn0x8PROxKeG8e3zv7N3YwFRO4ox39CXsJ7N+g/7ZetyLZ+Gp9C45Y1xvHmrO5y0NM/Cqi92M2xiKjWVVnL3lvPL57tw2t0iOXdvGY7/ui1JsTf1x5waDvY6+GAiZLlN/ACku6O1a52h1DpDiDDkUmRL4YtfbvSquvyTnSz/pMEdSLMqTPl3H/Jt3akQKfSeeBIiNA76TgYhCDS5//Zl+Ra+PaMf1tRghqVGsXZfMVe+s46rRybz8Tr36rvdrTrOs5hd+xYheSOslocda5lgWkqxLYWVlf9q8X6u/36/z/If9l8DQLCumGtipyCE1KwJP05FOgFRb+R4rkuTtlZnN4qtLwJg1m2hzjmARPM1WCUsdQ7CYr+aVcDL087jtBdWsLz6fAKA/fUfucetU7jBeR4A+qowKqpOo4LTiDE+ilWmuYQGQJXjfKoc57v29SKXGONTGEQ2FfYrqXRcjD7SROSF3TEmBKFf8zj89rrf+yGEk6jJqfD9S4QbP/M6FmZ9jSo0sdEgNACX0DCIQ4QZPqXE9hB6Cgk1fE2Z/eYm56hYcoiKJd4rKUdO7kbdwQoMMYHIOgfGuCCCBvlPka9Q/B1o1rIhhLinfvM54HE0v4sG9GiLsXWSUg46aiNsJ1yWjV1hFOSn8c3g0wmqLuUfuzQnSHN+JqaSfMInXM+N/7ywifXizSsHc04/TSjMmzeP5b/vYr61r89zzbttFAM7RQBgdzjp6pGRszXWjRqrA7NBh04n2J5Tznd/5PDgOT1ZtegA2+b5fuA04OkLIaUk+yHNBB3YL4boK3v5a+bF57ctZ3SoW6dWhZtZetD90Zh4xwBS+kTzw+tb0Bt09B2TxHcvayvmdh8RT1xyGKu/3NOk32ETOjP83C4s/2Qn6atz/J5fAJMitDfmfXUOCm2SESHeutkhJVJ8RIyhApNuHyZd0/M5AuOYvf8tv+eJ0GdR5ujo93gDtyVcAMknwWDkSLQAACAASURBVPXuz8OqL3a7xBTA4NHB/L662qtdeIIJc82vFJS7Xyr0oRUUjunKQ3svx1Dufoi9XvEIWIYSc3Is2Ttf4XbDd3zhOJWKkutAmtFjxeGRFTZIV8KkyMf5vPhlr3MGWvI5af0TAKSeXcD+RXHIAEGvSdmIRkYhKSG77gcASkQW/c1NH7atQUojOXUfI2necuZJovlK9KKxP3kL6E1w+0ZY+gTYa+GSj9mSXcF5r6+hA0W8ZnqFVJHHFmcaY/VbuM36f8T0uoCbaw2YIwKo3VmMLsBA3J2D0Zn0UJwBy56E7d96XIseIRxIaaLWOZhi27RWDU0XaiCwTyzS6iDyou7glDgtduoOlBMYlI7YPhc2fwIXvQ99L2zbdR8myidFAe1n2WhJbDQ8uVKALMDT1m1Fy2vxqJRyHSc4jcXGvEGnEVBTzvlLv6Q2qQuBB3ZiqKlCHxbFXe98hNMp6fKw20S+/5nxrn/cGTNmAPBB7TDX8V6JYeyod55ceu+ppMW6v3jPe30NWzK1+fCWxIbFaqf3o4u4fVxX7ju7B/1nLKKi1s6iu8bw+KxfGVXrP2S0Q7cILrjXe02Pun3lFH+2A2HWkzh1mJ+WIG0OnHUO9CEmr+mXBr4rs9F3bEf6jU1q4lAKUF5Yg8GkIzhce4PP2lXKkvfTOfO63syrn/q56ZVTMdRbBg6lFxMSGcCcx5t+tMaGGgjXt/1Lcn6Z5tpz2fg/+X5VX6qbuCZrDAuZQ6Q+h26BqymydeaL4hdJDtpOdeRwirOrfba5LeECmLIMkoZgqbDy/v0t+xE0kGQUDA02AEvoYHoJzyS1T9iu5iPHmdibNUJKPjLOYoxes+40/Etbz5rFitDzcb74BSIrg4y0CwCIKdpK/z+biqwbz5jKq9Wvsr68B10Kquk56j50Qked00Knp8/AJCTsXwGmEHivmWmV2zZoppKfp8PuHyFlFLt6v8uS97UU8d3NOnoFan/nDdV2cmzagM+4byCJJXWEDo5HGBopH4cdnmw6dTWy9lXy8O9r0laeOr8vV41sNC1ir9OER3xv73KnE768Bkf6SqzObghqscuOlNlva/N5dVQQbngfk24HBlMZ4sH9YDj8KTVfSCk5/T+/sK/Q92e4gcm6lUwP/JIIRzHyrKeg82hEdFdwOmD7N6A3a4LI2Lr8PO1Jw/Nsc2YZyVFBxISYW2hxfJBfUUtMiJn0nAoKq2q5/gN3fqKkiEBOTotm2oTe7CmoxO6UDE2JbGJdP1yOC7HhMZjlwIVSytanjTzBaCw25g8ah6m2ggt/+oyalB4usWEIi+LOdz4CvBcq8xQJjcXGDaNTeWR8L5c4Wf/w6cSFuac2Plt3yOVQesdpXbn3LPe6H55U19lZubuQWz79ncggI6WNQvvOshjpZdVj8jPrdeub43y+yTSIh9hbB2BO9g4YctbacVRYKf1mD9YDFcTc2Jei/zYN311WYeOy/4zBFNDqPHEAOOxOZt++AsBnBMqS99PZtS7PtR8dZmS0n/+xBjFxXoRvweWUku/L/a/gmzoghvG39IfPLoXd9VaKnhPh0k9ACJw1NQi9HqcwYKtz4HA4+eCBNQCcGzmDSlsiFTKWA3VDKbEn03NEHGP2n8bbBZ+7zhEg4OSQbWy0xFDhSPSy0jRgFHuJM92FuORD6HM+O3IrqKy1uyN3gIynx6PXCa/P4ELTQ/TWHaR0bxB5GyNwCB0HwhJIK89BBMVQlziYXannE2cQdA3wnuqr27MI6/avEaYQjGlnYO4x3nWs8oc7qdMZCZ7xIiSnsenHAy7R1XVoHAPGxKH/djy2/pfR4Zz73J3aaqFgOwdKUljw+tYm99sA+Ptr9BvXkZLkAL78OYM+eQ4C+kci0kLYtmIumTKOLBnDkgfHI4CTZzXvL/T+tcMYnBLJpoMlnNQlhkCTnvIaG9e9v57fD/l2eu2VGMZzF/Xn5/R8rh6ZQmxoMw8tKbXQWYMJdi6ELXNg3y9Qp1lmHDKKQuuTOGQMkqZC3Beh+s+RmPjt3CkM3GrHvk/ry5gUQuipHQnsF4Mtz4IpMbh+CJIpH21kyY4CLhiUxOTBHRndLYaF23L5fEMmFw+KZ8YXa3CgQ4dkgflhEoUWbLjIMZRYUcZgXdvDtJcHncU4y8/ugvv3a07XR5DyGhuPf7edLVllPDy+FyO7RGPU63hpyW7+eVJn4sO0v80j8/7ks3WH/Pbz6IReXD8qRfPfqhdKf2aX88QP6a7MxQadYNqEXlxzcme/Vp/SaitTv9pKVLCRR8b3bjYn0M68CuwOSZfYYJ78YQfpOeUM6BTBtSd35rstOewpqMJs0PHN79mHdW+uH5XKIxN6/aXVqI8rsfF3oInYGDgOY10Vk3/6mJqUngQe3IXBUtmi2LDZbMycORNwi40f7zyFXolhrvp7Zp7jlYn0m9+zuGfuliZ9Neaa99bzy27/aU2mlgVSLpwMubQbez/XfDEiU0IpPVjJhNv607lfjM92npaK+LsHY4wPxlFtQ9bYKZm7C+sh3yaABVV2knQwMMiATAim012HtxLq67cso1OvKCb9n+/U5FJKLBVWSg9WwGfuFOdJ5onUOQeys+YJNlY7aJh6j9IL9AJMAiodkhC9YFiwJoIWltuof5GmY89Ieo5MoMfIlv1kKpctI+tW7Y01bdFP5M2ciTk1lfyd+awKnOSzzS3xk9GJ+nT2Zz2FbcUc8iv+3Zpb4hd9QhDxNw9A2pwIkw5rngV7pImyp92LMlfOuwl9THeMqWMxJg35S+cr2PgRa7u23uWp/2kdGZRYQOCA/uhCQ9n000FXdE1Uh2BOvbw78V3C0Xt8/u0OJ295RN60xMoAG+POSeXms7u7yiprbbyzaj8jU6P4OT2fGquDvklhTBqYRJhJBzqd34eH3eFkTUYxzyzcwc68Ssz2Oi7aswK7zsC8tFMwOu3MuX0si3YXM3fBBp4L3E/vcSNYkVlN/0VzCO7Zk8qvv2J+l9GMvehM+o4aSNnuDO7LjeC3fdpD7LyBHXh+Uiq6V4dS+ruVoFgL9v4jeErcTEBGAP/ir4UHC8MrpIsiPrWfya3GeQzXaQspSqkZmaTUIYTTtd8cUhr51D6BPBnAvcbPqXMOwKzbhhDO5ht6csMSpCkSWV2OrmQrMq4ftn2r0J9yJ9lltfxn8S7CAo2c0i2WBVtzGNM9luzSGj749QDFR2iRw6FiJ1MMC3nLPpGXjK+TrGs5JdRM2xU40TFZv4o7bLeTIVteguJIo9cJXr9iEON6xvFndgV1dgf3zd1CTnnLeWsig4zMmNQHh1Py2PztJEcHMbJLNMt2FvDfa4bSJSYYIQQWq52iSiu78is5q0/C8SU2hBDdgYuAZPBeMlRKef2RH1r70iA2jLtCyc/vyvcDx6KzVnPRwo+p6dyTwEO7MVRXYAiN4M7/fgJA6kMLXCbrBoGwe/duPvtMc0abWzsACybXsQax0VhM2BxOr5VU/YkNX0u+NxDrEFxbqX1h3frmONb/sJ8ewxOIiNcc3pqbn7UX15D3nNtsF/3P3hR/lO73XKA5wZkHxjJnxm+Mc2hfQonTRoBTog9rm+my9lAFRW9sQRdqIrBPNJHnd8VWYEEfZqJ84X6q1+ehD7LhsLjfIBLN16AXWsRF7qgP2VXQk1N2j2JHzenkBp3FqdcO5Z1Z7sS2I4P1xHtEqiyvdXLRkycRGOrfVC2tVpwWCwUv/IeyL7/0W2/FmJdw6rzfbvrbvyIl40ciu1aTtSoaERxHyJlP+e0j6alRlHy+k5o/i/3WOdKU2p1EGnTk25xe96aBBWU2v5aHsPJ9xBdsZE+3S3weD67OoTrYnVMkvGwvQ/7QHE07f/E5gQMGNGnzytI9rPghgwS7jj621lnI+p2axCmXdvcOq62n4OWXKX5zdpPyyCuuIP7hh5B2O9JmQx+q5RV0Wq1ULltOzl13tercLbE3PIm1iX3IDY5hX3giUzfOIa2iqS/Ss0OuQO90sCRFezl50qFnnN5tAZEOG5YVMzGmjEYX1gFDXO8mfRwusZdHU/flixDZheraMTgqWy8odolN9CUUm9RE30LdMs6XCVhl8+OLMMzmB1nFxYY/KLXdTq1zJNPYzwaCqPazdMKAjuEkhgeyLbuc7LIartIv5inj+62/0CPARmd3nrZdwe+yG63MpIDZoHOtn3Xvmd3pEBHIit2FfL8lhwGdIhiSHMnafcV8efNJrpWmW0tlrY1BTyzG7mOZjLZw8N8Tjx+xIYSYAHwNbEbLyrkBSEOLRlklpfT9ancC0Vhs/DDgVLDVcMmCD7Gk9nKLjeAw7nxPExO1Ngc9p2vm9gaBsHXrVr755htXvw1TKuBfbAA88NVWvqhf4v1wxMZdZQEY6/8BGk9H5D35FKWffkrSKy8Tdpbv1Ux9+WH4I3xiF0JHuxV/47ZJz4xuk+NZwewtWA+0fuVbZ95qOgxbhTVpIoGTbkY0ODlUF2mvciGxSLudr579nYJDVUy+fwiW0lpMX3ovMhd7c3+k3Yk5LaLJeGv+3M6Biy5q9ZhCJ19EzbY/+TX0PKpCkxn7yx3opBNhCiFk/H+86gaPCKF6neZQG9AjkogLumKIcL/ZSqek6IPt1O3WZi1DRiehC9BTd6jSVeaLup0/YO450atMF2ok/YwkfjxYwmXDOzE0KRxHtZ21Px9k6zLNgbVDtwhOGpvEbx+kk2+TPgXGhVOHUDntTuyb1/n8mnXojBzqdAb7Uyc2OdZ99+d0zPH+jCR/9CHBw4f7vZaMwipCjHpiQs3o9TrsNgeVxbV8/ewm6iy+JdAZ1/Wm+/B4Sj//gjkrYmjweB2z6l4MjubfCitDOrJh6EOu/a57vyK6JJ1gS34zraAgMIK4mjK+7DaWi/eswKEzIqQDnWyDFaA5jMFg8+VjIWiInNNFpBA89pEjc75jTIjhM8L1c7SIqXqsl3+FaY7v/0UpocY5CoPIx1Q/DSSlkRrnKAJ1KxFdRsP+lRCRQvmVCzn9rZ0UVdURgoWFN/YmOcIEkalQdhBqy+Gd+gwJ/S6Gbf5fMAAtNPzaBWAKxppXxf59pQR/d8B12Nw1gpgb+raLE66UEovVwXOLdvHBrwcY3jmK+//RgwXbcsksqSExPICPfzvos+3xJjY2AV9JKZ8RQlSiJePKQcvguVZK+Z9mOzgBaCo2xiDtdVz6w/tYUnsTmLkHQ1U5+sBg7vrAnQegpNqK3el05ZvIzMzk3XffdR33FBsfrz1Ax6ggxvVoGgZXWFnHsJlL0Iekc96oPP4z7oUmdbo8tABfInZ45ygm5kBlQQ2XPzqCqA7ut6Ly774j5/4HAAg6aSQp7/t+G2hJbAQNiUUYDUi7k6iLunsda9w2/NwuhI5qnflROiXZD7fembJynncIaPy0aURddaVXWeErr1L0xhsAdP7qKwL79nGdq2zeXqrX53nVRwcdn9bCfqWUVC1fQdatt3pVCRk3jk5vvkHJZ5+R/8STpM6fR0CPHkinE3tBAcYEbfn6rDvuoHLxEq1bHw+Bynk30fCQiLn1VsIvvABjXBzC5LawSCkpeO55St57jw7PPUv4uefirKkBIahYtoqCp17BUZlNp9kfkX333RiShuAo2IGjeDcpX87HGN8RQ3Qg4jCcaC0VVpZ8kE7/cR3p3C+G/VsKSeoeiSlQe+sq+/prch+Zhj4mhs6fz0FabeiCg8Fhx15YSNYvW9kfNIg9GwsYODSQqOc1o2fsXXdiLyyi9FPvlQMSZ84kfNK52EtKEHo9hhjfU32N+XNlNuZAAz+/25qcgDCidjGpo9OwFxZS8qE2DWo1hrB6VPPTWj0HhXPavwYjhEA6HEiHA53H38phc6I36tiyLJPVc90RTz16m4j69HECy7O8+uv01mxCTtWWg7IeOED+M7Oo+sWVP5FacwQHUv5B14xvMTjqAOi68heM9mzsa2fzYfD1BO77iX/EBLB3Twx/7g6hd+EiygZOotOG79Hl/QE2C0EXXsGPJaO8PPpNAqwSguu1eZpZR6pZ892pdEhybE4y6pyuacYGLp02jGC9QFTWoYs0k3ewmvJv9xBgc1LkkBRanRh1gn6BemxSsrHaQZVT0mFcR/YtzSJYBw0Gk1iD4GSPyLFDVicWp6RnIx8is24TIDCKTBwykgD9ekxiL/nWppaqljAmhRB+Tiqm1DB0Ppwqm43KkRKs1VB2CMePM7Hv24vFcSrB+qUYxV6KbI9T52x+qtKYFIJOVBDeYSO6qDhKN4QRcekwDPoSHNVO9OYaREWmFk0VkQyrX9T8SlJO1kLqQxNBCE1QbMinfPFBggbFETHeI6uwrRbyt8N/T4MeE2DsA5BYbz0sy4TFj8Jp03wmJDyufDaEEFVAfynlPiFECTBGSvmnEKIfsEBKecKn0mssNhb2H4PdWcdl372HpUsfAjL3Yqwqw6BzcufsdyDc98N0+fLl/FL/5WE2m3nooYd81vPF2yszeHW/lmfg4phP+DPLwpPn9aVHQqiXFaUxj07sTcq+WjK3F3Ptv0e7yi2//87BK7wfxKnfzafguedJfudtr3Jpd1KzvZiSOW6fiLpdCxGmEOrSvyFoSH9SPvAtVBqnqzanhRM7pX+rrtlTqPiavqnd/BGJ3RZQWPsk1p3f4yja1aSPyKuvJvrGG8h7bAYhp59G3vRHvY732rnDa9+aVUnBa394lcXd3h9DjJnse++jaulSrdAYTOw9DxHYvzOm5F7IWgc16cUYogMJHhrv95rqMjKo3rAFyx/u7KOxt3Rl39lHNwV7yOmn0+n1147qOY4EDZY2Xxg6JNJ5zhys+/cjjEbynppJwvTpBA3WousdVVUIkwlhNCKEwGmxsGPIMP7sfQNFsd4+P5OvjGbxzxYqCpum1vfF8AkpRHcKo+BgBZt+dL8Fnvt/A0juHY3TKflj8SFKcqoJiw1kww/Nh5m7rsmo41+vnIp0SoRO+HywOZ1O3rx1hVfZmdf3plPvKIqzqxFAbbWN3H3lbFmS2arz+qJ3+nskFGzCrjdTGtEdmymUnT2ubLlhO9BYiLQ3yytsVDghVAcTLu5KzSLflgB/6MknwvhfAvVrccoAim3TqHP69kPzhU6WEKRfRrV1LM7SnehyP6RinyAsxULQaXdQXdp0OYsg/WIiDa94WYIApBQ4CQUkOirr/XbAJtMwigzEsOuh/6Xw3tmIxyuOK7GRC5wupUwXQmwHHpFSzhNCDAJWSil9LaZ2QuESG7tDyM/rxo/9T8HqtHHVj79R2QkCsjIwVpaiF07u6rkGHivz6WnlackYNGgQ5513XqvHcOV7y9mq/z8AanMvwFamfbgOzJpAVqmF0f/Wkl0NT43i5csGUl1np9bmpE+HMBa+sZWqsjoufcRtmj74z2uwrF/f9ERA/COPEHrWmQiDAUO0FjpY9vU3ENCfqlW5QFMrQtpPP2Lq3Nlnf74sIxGT0jDEBqIPN2OMa5qV0TPPR/Q/exPYWxtH5bSrKa+5gMqfZ4LN4tXG2KkjXRcvJvuee6hY+GOTPv1hSksjbcEPrn2nxYaUUDZvHTXbtP+Byvk3Q735Wx/bi6BRd/vtz9gplPjb/H+ReN6PDk+crOVtAKTNhqO8nKK33qb0449bPf7miL37bmJuajnpV1vIqrWSZDa22QRcbrNj0AmC9c0ntsud/mizfjC+MPfqRd2OHc3WsZlD6fP7by4HVCkla77e6/cBHZ0Uwvhb+xEW3dRXoOBgBV8+s9FHK9/odILrnhtNRVFNi+3C4wK54N7BBIeb2bI002femdYQEmmmqrTO7/Exq+5FCj1Ge/PhrgDVgXEY7TWYbJpDuETwZ58bKIxtmkYpMXctafu+JStpLDqnjZDqHKKL/wQENmMwFWGdyehyHl32fUdk2R7shgBAIKSTXac/QlFNMGHRZvrueI+AnF3kBnRjR08tk61RQDezjiiDIEIvcALG+s9heo2DGqekRkKx3fvZZRbgkZ+NUB3oBJiEYGSwHt0Rms4IHp7gso4aE4OJubEfWAqwLPkWZ/4BdLo6dOmfU1dhpHRvKHQ4H31kKob4egurtRphal1UUmOkrQbLqucIGHAF+uiuTStYc8HUtsSQnf495rgSG/OAhVLKt4UQz6KtxPoRcAFaunLfjgAnEE3ERr9TqMPOtUvrKIv7nYDsDIwV2nz5vb1WwbULofOoJv14ig1f+76osFYwao53X7X5E7CVaKb9vTPPYWdeJRNfXc1bVw/h7D4JXnUdDiezb1tBQIiRG553ZwFt7Rd6r507KPnkU/KfesrlY1C7bS62jCUEjRyJ5TdtUbHEmTMxdUnFmJCAo7ycgJ49XX04LTacVgd5szb4PEf01b0wdQ5HF2hwOfPZy2rJm7WBgJ5RxFzbp/5m5FB130AyVzbNnRB13XXEP3C/+5xWKwcmT6ZuT9OQvR6bfyf/mVmUzZ3rKkv9bj4B3b2ngHKnP4rDdqZrv3bLZ5h7ndfqL4PYWzRTZeGbWjRR1JU9kXZJ6ReaBcZTaDTG8vtmZG0NIiCA/JlPU7t9Ox2ee46cqVOJvetOYm6+GeuhQzhraymfP5+yz7/A3L07nedoPkPS4UC08FBvjN0pya6z8tqhAiwOJ/+ICefcuAiklCSu2NKkfu7YAV6CI7PWSqcAb6faczftYUOF98Ns/5j+mHWiVV/wtuxszVq9fx+ZU9oumoJHjaLT2281ey+Ks6tY/H46XQfHMuisFMryLVQU1ZA6INZvG4D01TkeGWS9ueDewcQmh7J3UwFpg2JdU02eOB1O3mxDpM3Nr45FItm2PJtfv2n6uY7uGEJq/xgGnN6JgGBvp+Qdv+ZgrXGQOiCG0OgAbepHSnb26YsuIIDumzb6FY/SZsNRWcmha64h8p//JKB7dyoXLybikkvQh4VBYCB1dQK5ZxuHrrwKgA7/nkXYpEnYCwuxrFtPzlQtxX/nr74ioHs3pN3OgcuvoG5XU2skOp2Wp8QHVcEdKIruy74uvl/UwikloiKD7ucOIXlgB2wxndixrpABp3WkpryGFW/+RnaB9rfofPBHUvcvQNRPXTrCkhGRKThKD2GoKcQZ2QWDpQhH9wmEJ7stBzYpWVlpxy7BJt0Jpq568iTCYwORNhsVi1eTc///gd1/OH1rCRg4Gl3kWPThyTgrN2Ls0AdHpbcArl76GM7KXHeB0BM0bhr6sL8WMXO8iY0uQIiUcqsQIgh4ARiFtrjaPfUpxk9oGouNn/qNpgYn1y+1Uhq3kYDsfRgrtOiGe3utgqu/hbSmZvG5c+eSnu6eCmiN2Oj3Yb8mZY6aJCwH7gBg47QzuPPzzazZW8xrVwxiYn/vlUMLD1Uytz700dM5NOuuu6nbtYu0Hxdy6PobqP71V5/nT3rpRbLvavoW3/nrrwjs04fy777j08+/pXNuFqm57jnonunbKf92HkHDhmJK1mbSqtbmUDY/o9nrTXhgGI4Kq+sBHTOlHwFpEUhLGXUPpbF/kdunJW3BfDImnEfizJlETG6aWbFm258cuPhi1374eeeROOsZzcReV8fBq66mdps7lXn4RZPp8JQWFbLl/Asx7NqJOXk0AYOubnbMmZencUFJAZOkicfDo3C04k00/r6hGGNaTnw0dv1OdlbXNnmwHyme2JvDG5kFLVdsA4+ldeDxDP9ZXhvQAdltvC5bbi7GRPfbWd2ePRy47HKc1dV0fOMNgkeOQAQEUPPHFgIHDnA7CB8lrLV23rlrJeBOPNeW7JtOpyR7ZymJ3cLR6XUs+2gHu9fl4fnVe+m04cR0bH2G1RMRR1UVtVu3cuj6G1xlUddcgz4mGp3ZTOTll+Oss1L+9VdUrVqNsWMSsXfcgSMoHJNZh6O6Gp3Z7OUz01aklFh++w17YaHLl605qoI7sH5Yy863IVVZJFTuoNZhpGvGtzh1BgxmHeFnn03C4zPQmUw4rVbX9J+r/9VrMHdNc/l8eWIvqcWWV+2y+IJ2D3XBwTirq6lNT8dRUooICcaU3B9bbhGVyw7hKMkheLiZoIH9cZSXU7fvAMGjzsCyeQ25992HLiSUjm+9i6OsCF1ACKGjh7eL2EBK2ewPWu6d8UB0S3VP5B80rz2fPxMnTpRTb7tXPn/JBHnRkH5+6wHyP//5j3zsscfkY489JhMTE/3WmzJlimwgbUZas32e+9iHMuWBH2TKAz/I8y692m+9bim9XX06amqa7XNGfIJM79FTpvfoKWfEJzRbd9ruTBm/bLOMX7ZZ9jabm70me1mtzHxgpfzprg+b7XPBNe/IzAdWyswHVsobT+3st97gwYOlJ831+dZbb7nqzZ49u9m6+Xv2yvhlm+XAuT/J7lGRzV7T3uoaGb9ss4ya/Wmrr+nqUyY3e021Docssdpk/LLNrb6mt956q9m6TqdTWuwOOTe3WCb08f85DZxwoevv2dI1Rc3+1FU3cMKFfuvF9+4rH9uTJRcVlh3Ra/Jk8ODBfutdfO318s2D+TKv1io3btzYbJ8bN2509TllypQj/tlr7TXZrPZmr8nzO+JEuaaW/k5TpkyRjpoamf/883L1J58c19eUMfFc13dkc997J/ecIF+7aal87aal8v4L32y2z9emfSFtdfYjek2XnXK36/wP3PJUq/9Ogwa5/k4bZTs8Y1v0xpFS2oUQ3wA90RZP+1uiDw6BlnPCUF7etvUc1mSvabHOhgOlmBM0ddtcJsPgSO2YNSubjDPOaNM4muOdLPdS33aDAep8zxHbCgoofOU5hLOCym9bnr6p00F58CrqCpq/sX1W/0mxzc4LPTo1W+/z3GIeXa45flp2ZzVb9+6vf4Tho8mNiac4OgFK/IeUjlrn24zemI8GpPN8vpb91dwjEvwE+GytrCHll6YZNX1x365MepdWMjqyZbcoz2mQYqt/0+75cRF8PHYAeiHYFOaguVead/t2xpaWTE6tjblh+RZl6AAAIABJREFUQfi2jUFSgIkZXTVzbt64ga3MQnDk+KGwjJUZOczIyMG224fZ3gfryqqw+jHlg/YtfDQxGNs2Bfa/gi4ggLh77yVz06ZjPZRm6fL9d1QuW07+M8+gKyry+73X86QEBp+dzP4tRZQ5m0/Odii9hLf+Twsg2L7Kv2VQ1ocd7t9SyPpWOiIDHGwhT8+ar/eyd1M+VSV1FGX6Wa/hKNHaaZR1aE6hS47+kI4NDdMopt3B5OV1Z3Gfk8mIiWPKrp/Q59VitwoiM7Spigt77Sd+8rME9ZsAM8K1Dv7vD4hKdU2bDBo0iIyMDO655x4/Z9TwNYXiib2qBzWZ17n2G3JwlBda+GT6bwSFmbBUaBn3bpt9GtJmY2c/70iQxtEYAJm33oaxYxKlH33M5u69qTrvAm688iLylixjcHCHJvV9cdZvK3nowzdbrGfoMISwM4ZStSEfU2wMImo4H3c28nIP73/MuQ/eSmx5KWETJ5L0/HOu8oTl3pEjQ8KCWPD/7J13eFRV+sc/506fTHoIqSSBQBIg9I6AYEGxYFtddRWsYMWfvYu6ru6qq2tF3dW1Ye8rKiJNVEpA6b0kkN6TSabf8/vjJjOZzCQEV13Xnc/zwDNz7rnn3juTuee973nf7ztyAC0+H4/ur+x2icDcWoTTGphOL/9gIc+f3kV5+U7MMf2L4TmzmbtDM7ZKjx7K0vJtXLDT0+U+cRV3o/ccoibzxR4dA+CLUQM48/s92H1dT3xvD+3H5ITwBse124t5uyK8sXR3vzTmZPZC9x8quHXsup1ssYdmg8zJ6MU9uWlstTvwSMmImB8XMHf1tmLerex5FQVL06c4Yg5f7LCdPw/IQCcEdR4vczN7YTzMko1X9bKqdBWrSldxWeFl9I4KzVryql70yn8u6+K/nf2N+9lVv4vxaePRCz1mvRmlcxXB/zBSSoq31KI36qg52Iy9zkV6fjylO+vZ+NWPzybqSFKmjdEzcuidE0Nrk5u68hZ+WFJCzUE7qq8WT8tiEAp682iQLlRfHXrzOITQIVUHoGKLM3HRw9N/VTEbJwIPAfcA64GgaDApZV24/f6bCDU2xrM3SbtRzF3xIR63QsLetUjAXjCK1Hgrc+bdHDA25m2E+Gzmz59PWloaGRkZbN68mVtu6X5dsPDlQvrUD6TRXE2jRXvCz43LZU+DFhzmc6bgbR6IVE146qb4jY2n54bWg7hqwTQ8ZWXsmXaMvy1v4w8opmBvyPrGFk7asJuiUf1pHj2Kqc++AWjBgC+V1nD77lCd/qHqajYq40La39u4isJBeZRec2231+lHZ6Ti3Gc4d2L4yWVdroI+OZ+zf9jL84OyuX9vGUvrQi3wPZMKyf16c5gRek5c4150TX+hNlMrSjZOLmNL/QHsCZpxZ7Yvw2nTRH4uTzXxwRpNLVNVbDQnXo6laREZCaPwRB3FLnfPJ8vommcxuPeA9PL+jH+Qn5CL06dyXNFOTu4Vx5e1Tf4J2qwInB3EVeZm9vJ7EL6qbeL8Tfv8267L6s3G5lZi9DqeG5T94z+YI0CV6mFv9E8UV/KnfeXd9gEYFxvFmNgobs5J5VBzMesr1zM+bTxptjR8UvJKWS237TrE7PQkru2TjA8Y/Z0WHxVd8yRfHHcHfaL7sKvVzbSiXd0frBuE2oJUwn+fZyXpeWRgAeYOwaif7f+Mm1ZqgcvhTLo/HfUnTup7Es9teo5nfngmaNuNo25k1qBZAGyzO5DAIJtFcz0Dm+0O7t1Txo3ZKYyNi2Jzs4Mh0RYUIahxezEpgmj94T0kPtWHTvnv9aSU2cuY/l43xf+Akb1H8tCkh0iJSsGjemhyNZFo+emK9P1USCmpLmkmIS3K791qqnXwxQtbGXtqDvZ6F8te1bypJquelBwDo04eiN6gkJBqZcuyL9m89Av0JhO9c3Kx19dRV3qQ+vIyDCYTXo8Hj7Nn6d4AN7796a/K2Oj4yNVxB4G2DvTf+1fcht/Y2B1FRXmwsTF6/zaG7t1DVPU+XGnZgJZSdfc98wPGxjUbILEfDz30EEOGDMFgMLBmzRruvLP7EtSFLxcy9zutDHj1hSt5b/d7XJ37f3y07jMOxu/AVTMVU9IyAJ6fvIjxOdpSQmdjIyrOxKmDD1D5x4Aktql/Ln0/+STkmB09BW/ccQ3nPvBkSJ90k4FSV+AJPqnkAmr6hKZqDo+28tmoAWwfNBh8AQkhW5oTe1moS7E6LoGzH3za//741StZPG5ySL/OnJeawMLyrm3aZKOeTRMHs7RkKfOWzWN82gQ+1l/R7Zg/jB9ISofMCo/qoaa1hglfvYwj5oSgvr1KQgNIhycP55UTX/G/7+yBAXhpcDZjo3UsKf6M03LPwKDoKG8pD7pxLpyxkMFJwUqDqpQIQAjBpVv286/qwPLc8YkxbLY7KO/w/dzRN5VrsrrW/vh3Gfv6WFq9WhryrWNu5aG1D/m3KUKh6A9FGDpItm+p2UIvSy921u/kzlV38sej/sjkjMnUebzct6eMNyt+uueTuIq7MbhDXc0SHaKDrFVjn5dwo8dW+3cEXgzO7XhMubTE/Z7Y6sfRe4Lj3O1x54f8HbRjcG7jnVFjKIhNYcz7s2nsHfidJx66GkUNXk5VhRWvMYumpGux2L/Cp++NtfEDdL4aPMYc3OahOGJP+VHXn2jQke5cTnmppp2Tn5DPgmOfR5VeWrwujluzAdn4JabWtTiij+PlsWcxKa3n+g8/lsUHFtM3ti9u1U2iOZFka3KXQbV7G/Zy2ken/STHVYWZxuSb0HkqmaIv4qljniXaYPlFlDx/LPb6Opz2ZvZtWMfXC/9J9tAR1JQcwF7/434neoOR3DHjscbEsuGzj0nqk43BZKJ8d+gS46/N2JjS3XYp5Yrutv83EGpsjGNvUiBC+KJti7AedOBrU7oTSO658Wp4RMt1rjv/S4rtBj755BMmTJiAXq9n+fLl3H333SgdXK9SlSDwp6UNeXkIc1drxsYFT4zije1voH+lAGeLlwXj5wWdo0VvYe35a6kts/PmfcH6GVPP7YeYE7gx9rr+epIuvyzstbZPikNsFjaFcXFDIOXRpaqMfHUEAh9SGGiJOR1r08e8OnM5Mzbs5pR4Hzt33sRT4x5E/GEWZouH5OxmbL3cSFVQv9vKpvEDOX/Ew7y24yn+kH+1/xhXvfMKZy5dxJwH/sjuhDA54x0oO3ooJ6zfxbhYG88fCsR4mJu/It6xEq9zH3+Z/BdubnvC/O7c71AUCw/s/J4784Zi0Rtw+3ycv3k/zwzMopex60qNUkrGfbeZYpdmYycdvBghtYn9pekvMbL3cJqat+BwlCBVLwkJEzAak2lVVWZu2I3TUULD/ltYfvaXJFm6VsTsuIR2/8T7OS03/M1WdpGWCjAjKZbnB2Wj/zeqPnY+lkSiCAW3z82VS65kTcWaIxojw5bBIXvXMTPH9jmW47NP4Hu1kBExUby/5WG+qnfisozGaw5f8bgrTPavial7/rD93j75bQoSCwDNG+NVvRh1AUOzzF6GQGD32LEZbBz/XnA2v0RgT7jI7+n6rTDUsZDHxs6mIKHAf09qn5Tf3/0+UYYopmd37VEobiqmqKKInNgcVHMeH1dW8kFVK6PUxawp+ReKr9rv7ZHChBQmkG4UefiiYp05e8DZnN7/dAYnDQ5qd6sq75YeYP7u/TSJ7j0Z8Xr4YlQBmWYjqxtbMCsKeVFmaj1eDrS6yLGaSDMZfhJNjva59eDWTXyzfj335Yyk3ieJUgQtnaSgFZ8XVdeWrluyizM+fx2d6gsZs53J519Ec10Nm5cuJnvIcFL6DcDZYic5uy/JOf1ITM/UMm+8bp4sLufxg6HVjXWecs5PS+bhwlG/HmPjf4HDGRsX7PiCmGK739hASubfMBf+qmlNzCeQOtqvby59++Xw5Zdfctttt2HqsIzx9Nyl5I5KZvqlg2n1tDLx1UlculaLT2hPW233Wrww9gZ8ihboN7CskBFlJ3L9o+fy5n1raaoJ/rGed3Q1FR3SbMPFabQT7gm8MxVTtScfj8/Dee8OxSQkGx2BdeabRt3ErWVpqHptMr1WPsJYAiXQ83VTSV/2DhJInRJqi96Xm8YJf5rP7t99AcBWBvMncW/Yc3m2r5e7l1/EWQPOYmrmVM7eHePfllRyAQKI06m0qgK31G4Sm2f9e0ssDp9KzspNTIyzMTtmO7evup0MWwafnfkZy1cMwefrXiTpqIlrMJkOL73d0eB495R3yUvoerKVUnLtjhLeaYvP+H1KAo/lZ/5kT2xun5uRr2nSy+cXnM/mms1sqg4Esqbb0im1B5bYzux/Jufmn8tZn/S8hkxP+PPxH/FxncLK+mZaGr5BSBeW5s/RewIGjD3uXI6JbuSmYReTbkvHZrSxrmId+Qn57G3YS5wpDoPOwAnvncCVw67kiqHde7mOBK8qyVgRbPw9XdCHGb3iuHHnQd7rJoYkzWSgzBU+5ueLEblsPPQ+NkM02+q2sr12OzGmGFaVasJ3Vw+7mpe3vkyDaqIu7WH6tb5NfcM6nFGTcMRM73Lp58cQV34Xiq8endqIRMfRg/5IrGM1O+q3s6OxEpBIxUxT0tV4jTmHHS8E1Q2ooAR7P8fFRjE9Ho6KdjM4cSA+6UMndEF/415Vcv6mfayo7zrA8cbsZB458OPTvW06hdOS43mtPBBw+WphDl4pGR1rI8kYuBfWur3E6HUY2vWDPB5KDh3krNVbKEv58eLac8w+5gwewNJmN1lmI0fF2xBC0OT1MXjVFtw/0dxdOW14xNj4JelsbLw48STc+sDT7/k7FhNXbMdnDngp5s+bDX8bSq2i8KQa7IWYMWMGixYt4sYbb8RqjWLFGzsZc3IO/7xFyz65asE03tn1Do+seJwL198PQObABGbMLfRHK/+QupQtqSuxm+r9Sy3WfAetOzTthtEnZbPu0wMATFt+lf/Yurg4BqwOTPwdaZ9EO7NgYBZzt2nyvEXjB5JhNnLMO8dQ1VrF45ma+zx9yEf87tNAufHqTssqcbKOO5hPGtqEdMzEjVQ/fxyFg4ODSB/Ny+T8tETq69ay4YfAeOlZ17DKdD4XpiUigT/tK+eqjASOeSu4YJdPF09j8i3EVD/GU6mh7nOvIY1JI1/DZEpBpzuyCrQAh0oXUlr6OhmF7xBnNAWpYR44sIC9+x7uZu9Qxo39gqiorj03o14bhautDsY/T/gnI3uPxO1zo1f0IfEQUkqu2FbMvKzeFNhCNTycXidnfnwmJc0laCuegj9P+jP94vrx0d6PeHWb9p0NThxMmi2NBHMCb+58s9vznzdiHpcMvqRbo2bh9oVY9Bae2/QcpfZSFp2xiOUHlzOi9wgGJQ5CSkmts5azPj6LWmdoxHxHz0NHPD4PK0tXsqtuF6NSRpFuS0cndGEDL39pnF4nEolFH/w93LrrEIuqG3h9SF9cqiTFZCDDHKoNoUrJ8rpmpiREH3EAr5SSgwcP0tDQ4C/82GowUW+NJq2xhqlHH82QkUOwWW1BHpz2fbfUbOGM5U/SnHRluOH/LaLqX6Ml/g8/+biH49o+yVyd1ZuYDjEsUkr2NOzB6XVy3bLrKBF9aU66qptRes5Lg7OZs/UA7h5OoUd/u4ghO9azevgU+rr38nXiShzWbNymNEBB4MXU8i3NiVfgigqNj/uxGJzbOC/JTWHiIOobV7Ps4DLuGnc3Jd54rt3wGSWnXBExNn5J2o0N024r5eV5LJgS7NI+avdGhu7bG2xsXHUuPD2Gry1mvnIEnpzycweRNyiXjz76iHnz5lG2pZXlrwevlV21YBpPbHiCt9d+yLk/BNZ7R56YFVSXAaBYfw1Z3tC4iotuysEbl0LJpLHoVO1pKfGKufS66iqEPny0+95WJxPX7GBqQjTL2gIv14wrIMti4q3yOsbHRdHHok3QhS8Xcmy0h5PjtLHz8x5gh68X1y3TSnA3Jc7BFXVUyDEukQuYxpfEZ8xlRulxQdsGmx0sGT8egGXLC1BVd9D2mOghjB79AcXFz2E0JnH/1q/4ujQ0h/ShSQ8xMaE36zecE/Y628nOupLs7CuRUkWvD37y83jqAQWDIZay8nfZvr3rYN7Roz/CYs5k5dcjAOjX9ybS03+P3b6L1ta9OBwHKS55rsv9j56yFYejGJst1HPR6mll7MLQugcAn5z2Cdmx2d1eYzt2t53xb2if7XkJLsZEaW7YP5abMQko9SgYhWSAyccWp46uymQ/OuVRblhxAwD3TbiP0/uf3qPjHwk+1cfyQ8t5ftPz/OP4f2Az/rKCVqrq5uDBl7Dbd2E2p+L12amq+oKRIxZitf6IJ/UesG7d6ZjMaRQOfuqw3qjy8nLeeustGhpC3d9dMWjQIMrKyqivD3hWpk2bxuTJk/H5fLhcLqxWrWyAlJLq6mrWNDsYGm3l1ZdewONwowLrs/NZn5XfxVGCGRJt4fUhfbtclix2uJBAltnov2YpJXfsLuWg082CQVmYFYUGj4+Nza2sa2zhxdIaGr1dLyEAnJ0Sz9/y+1Dm8pBsNPi9Cj3Bp/ooqiziu7LvmJQxiX6x/Zj1+Szy6pPo/XklEtif2Z91Q49i0M5vGLhnD5vzR7B4ypHFlPTd+x4NpuXo1MB3KCTIDqdq9BkxqAYcOgeqooIEgcBlLqCxt1ZXS+cpx2cIliA32VcSU/cC0P5IAYoumglpE0mzxHFyv5MpTBjE6tWr+frrr3G5XGRlZeFyuRg5ciSffhqoIH7vvff+9owNIUQC8A/geKAGuE1KuTBMv9lt/ToGFJwspVzek3GEEMcATwN9gDXAbCllt1V1emJsDC7dF1QP5bZLZmD6x3GssphZ0sHYsO7fxtS51/HpF4u58sor+eGTanavCy5VPfepo3n6miXoZLBRoHrLELpEhAg8kUc1baclJvSpb9rKa1BsUahNAXdi/rat3SoqXrx5P4tqGnmlMIcLN2tegfYlk440u5uZ8MYEv1ejnXFjlzDu3VMBeP+Utzn/hx2UiNCb86vyLC4Q7/rf58qd7BF5zJTv8QfjEtzugIvzmGl7+WppoBqhzVaA3a4tA91XZqbOp/DuKe8y67MLGJ86ilOVFRQOfoqNmy7t8jrDcfSULeh0lqBj/RhstgLGjvlXl9tdrkpWfTMh7DaDIYFxYz/HaAxeW/apPoa9Gj5ob93561ClymkfnUZ5SzkbLtiAXuhpcjfRUv81ipD0Tp7BsQuHcEa8m3xz+BTa71t1DLcG38TbP9+/Tf0bqZYU2FWBNaOYlNRTMOhjURQDdvtOjMZeGI0J3X0sPwsNjetZv/7ssNsyMi4gq8/lmM1aqrbH00RjYxFOZxmVVYsYNPARdDobVVWLqaz8GL3BRnKv6dTXr6as/O2wYwJYLH3Iyb6a5OQT0elCa/q08+ot86g6sJcZ19xIwVFH+9t9Xi9CCJQ2j5jTWcY3304K2nfvnvPIzZ3KypUrD/sZxMRUYTK10NyciNMZE7J91qxZ5OQEfoNVVVU888wzIf0yMjeTk/MDPp+OfXtH43abcbut2O1dxzk0WKKIdraikxIVcBhNRLld3HTTTURF/XTLNl3hk5JtdgeLqhu5OCOJJIP+J1k2rDqwj3fuvwNFp8MWn0jVge5Vjzvz/aAxLJl0KqcufoO8fVvxKjoMU49j0qlnkxUXhcfTiE5ReHDNgxydeTR9bX25bNFljKsaR4LryH5H3yd+T4u+hXhXPI3GRurMdaS0plDQUIAp3kSMM4bW5sB9OiUlhYqKim5GDOa3amy8gaZgfAkwDPgUmCCl3Nqp32zgUill6GPzYcYRQiQBe4FLgU+A+4FJUspu/VI9MjbKgl322/ou4vdVh7inVyKnFJ+CUdXclda9W+g/5Rg2Hqrksssu48P7eyYyJKUHV4PmwTBEzURn7HpSzCr+gn77Pw5pL9ixHXweUPRhC8W1x2u0ezO6Yl3FOt757g9Mjw0Vh+qb/whWay7Vh/5OZdW/QJXEFjk5a+CNOKNDJdzHiyJuilrC4mYrY/kWI4E1a6s1h/HjlmC372TN2hldno/V2o/W1vA3hLwB95GYOAmLRVsf9flclJYtZPfuPwHBE29GxoUcOvRKmFECGAyJrF17DMOHvx32uEdP2dbj5ZmKyk/YuvW6oLa4uDGMHPGG/31DQwOPP/44AIrFzoepn+ORgrtSHaxr0fF5U6gLPsYYw229KrB0k3Wal3c/O3fe1aPzBKjbGUtCXteidAP6/B1ciTRWVTLo6GNZ+fpLpPbPIyE1HXN0DLb48DfRioqP2botENOUlHQsmZmzSYgfj9NZhstVycZNl7V5mjTS08+jtDTkOeRnxWxOx+kMTfsGSE2ZQ3mF5rnav2citatdWGig94gakgZqT65ZaTcjW/vx/Zqbqd0ex4DTuq8aWlOTyfZtR3e5PSHKQkZGHbG9A0a7otjIz/uMhoZG8vPzu514S0pKePHFvzNpcvgKu13iO4NVS3VIkxZPYbXWEx1TQ2VFLl15w3StzVhKdhM/bgoHG7SHn2uuuYbExIAh43G7+PDP91GyZSO9snJISM9kwPiJ9B02mqaaaiwxMWxdvoScYaOwxsVhsWm6MqrPR315GaaoKKqL99O7by4ep5Py3Tv49AltSfPk627FEh1N5qAhwfEdHg+KTqFy7x6c9mb2rFvNpq/CV88G+P19D6M3GGhprMfrclG2aztpAwpIzx9EVFw89eWl7Ny2lZVFRfiam5gyrJDoPn0xxsaTnNyb+vp64uPjWbp0KZs3/3txYz8FKSkpTJw4kfT0dDZv3owQgqVLl3LiiScydOhQzGbzr6vEfMhOQljQaqPsPpzHoMM+UUA9MFhKuaut7VWgVEp5a6e+s+nC2DjcOEKIy9E8GRM69K8Bhkspu5SCPJyxMXHPJgpL9wW1vZfznv/1iSUzsPq0tVvr/m3kTz6GDQfLmT17Np8+FLwfBCKVO/4wnPV/DepjipuHEOGzijvGaACYYrUJvO/S1fBwP0jMhauLQgyOdmPjcHU4jl84iFtStCDUlJTT6Nf3Rr75Nqztx6jvG4gdcSPknUjBspuoTwvENChqCwenjqOosohrl1zCn9KDs1+mTd0TdB5H6nUY0P8uMjNnh91WX7+WDd+fG3ZbO/l5f6S+fjUGYwID+t+NlJL77rsPAJOpBZfLyqjRH2GxNBMbO4q+fa8jIX78EZ1jU9MmiktewBaVx779j/nbDfrLWbpU+zyMxlbGjnuvqyFo8sEH9Ua2OnVkGVWuSu660idoE/rQIc/h9Tj46JlLcXl303t4LR6Hjm2v5yIEDL5wF4rhyH7/ZWt70VJhxVFtRvUGWzrTr7iOgZOmouh0qKoPRdFRV/cN3/9w4REdozPJySdhsfRBp7OQmDCJmJghrN9wLg0N4Ssah0PK4J9C0bpTcThiGTRoEGeddRaqz8fWFUvoP340364e0/VAP5LSQ/ns2zcaIXwMzl5EXGbw8kjjwWgqizPolXqI0u96Y07w0GdSGcborgXkOpKTfQ3ZWdfgaK5n79YPObihGvOgv4f0O/RtMrEZKtaUBnTG7guIuVv0GKNC++xdV4DDnEpTUy8mTHzL397Q0Ju4OM2DW12dxZ7dY/F6DYCCzbGflL4l9MopR68PvaaNrw9CtatIoxFjgoqnHty2JIz1FRhMXkxxLlrKu/CkCElc32ayj9UMRWe9kV4xs/n67yuRCGLHC/oO2UZFeS52ewKVlX0xmVqxWJoYNHg5jeWpZPV5gOyCQezbv58hQ4bg9Xpxu90UFxdz8OBBpkzREjK/++47NmxYwtBhn2MyBbwJP3x/As3NSbQbY0K0exAFZnMzHo8Zr9dEbm4uEyZMICYmhtbWVtLS0tC3LXn7fD7NI6YoeL1ePB4PHo+HF154AYfDQXR0tH+JLCsrC5vNhtfrZcqUKaSmpmrBo01NFBUVMX78eCyW7usyuVxVuFwVxMYO/fUYG0KIfwJrpZTPCCGMaMJegwA3cLqU8rC1vtvK0X8rpbR0aLsRmCKlPKVT39loyyAOoA54FXiwTTq923GEEH8DjFLKKzps3wLcI6Xs8m6uWKNkvwEWRKuC222lLC44kyDW0UKUK3iitDZVU9xWMyzRlYgitZuvztECOhs+o6R3797UHQydGFSvpiKn6DMAgaW1ghZj8I/Q6vbhtGaH7BvXuAchg5/YrcnukH7EZkB8YH+nqrKhqRWrTmFYdHj3sESyq2YjyfrAOZtNqVgsfahvCJ8CGd/ggWzNEClpLqHUY0W2RZkbXHtABq7LoOjQ4cMoJD4pMBkTqHPWUdirELPOjKo6cTor2NVcjRDQxxi6JGDQx+Hxajfr+LgxdPW01X5FIKipOYBOH1jKiosdASghxlxLSwvV1aHy6Yriw2SykZwcrBVQVlaGoiikhCmkFHQWUuJ0OlGUJlodB/ztTY1a1dGY2B5o4YehxmukVfViEhKfMFKQMBCpCsp29UxiXSgSqQrM8S70Fh8CAwaRQX1ZBUIBU6wbn0dBb/KhhPkuHDVmfG7t714I7bNuv6XoTWBO1G7Iel00Nlse9padeL3hswhsUQOwtwTEuEzG3pgtmfi8KgaDASkl9fX1tNbV4NVr3h6dzovR2IqiqLS2xiBlZ1ePpE+fLBRFobW1laqqrjMUdK3aefkstiDLRK93E2Wrx+00I1QfBmvw79TjNqMoPnSdJ1EJzc29UNW2c5Kqdm9AGz4qNXiJsiuctSa8Lu3v1JbWs306oroVWmu6ltFWdBKhSFSfQADWlJ4LQv1W8XpM6HQehNL2N98eGPETIYSe2JjhCH8AuPajcXvqUYQenc5Gc/MWfKr2XVgtWRiNSQjRbpi04nZXYzAk4PXZ0ets6PXBCsOq6sHna0VKDzqdFZ3OitNZjtN5CNnB47tpo/MXMTZ6qpk7HXii7fWpQDSQAlw6WCVGAAAgAElEQVQMzAcOa2wANqCzj7axbazOrAQGA8VoRs1bgBd4sAfj2AitYhL2OG1ekMsBhKXrtVkIXychyklbzIVEdIz6kap/h8qqKgzEdju2Ld5EkzfU2m816hC+KoROs2gSfPvwOAwhhkY7W0xGHG03yWEuF4bGQ1QZjFiM0UQbo/1KlImGwNfuclX6Jz+9zobXZye5w/xrMWf618TjYkfT0BhcQt7a6gNLvP99siWZypbNSGEi2hiNXQZf19Dk4RRVFOFs++W2ODXRms3Vm0m3pdPsaabJ1aR1lhAbOxrVZ6fZvh2jIdGf1SGl1//D6x5BRUUFTifodIkoipeoqBSamlqxWCwY28q/19bW0twcmASTkpKoqakhPT2d0tJSVFWHw+GguFhz5CmKQlpaGm63ZuQdOHDAr1PQGZ1Oh6+D4FlUlBG9QdsvxMiQGSAOoVN6EROTQ2vrAdzuZiShE4BJn4a5rAFz2884OasvHoebquKuaym0n2NiRh+sMR3+LqXUvBFtuf5Rscn4vB6QoKoqUqo4nHtQCZ7sLEmH10tQvQk0NhmprT2IrlWld04hXunF5d6OTjETHV2IVCWKTodeN5Da2o7ZKmGknfWBZSXZ6sFt1yH1JhSvA6nTofi8JPXJRlF0NNfVULojsErb/qctdXpMcQk4HIHP1WcNLwfv9RpprogmPbc/iqJDdGO4pKSkoKoqbreb2LhYTPpmaspKET7NQ5DSbwCGtlR4n9eDUCQORzFebzOq9CBQgiaC2JgR6Dz1NFZrhrK9zIqiV9GZVFS3gs+joDP6sCSFPtDohIWY2ELKd+8C3CSkpqOqKs21Ndji43E7nbhaW1B9PuJ7pWFLSEL1+VB0OlyuKpzOUqJjBqMILfhTSi9OZxludy2q1P5+FWEkNnY4Pl8LdvsOrNZ+6PU2mpo2+/u0o6o6jMZojIZ4jMYEXC4vzc3NqGoFRtORG1GdcTiicbusmEwtmC32oG0mYxoud2gtEtWXBlSh6II9OHpDp8+zk6FhNqVhsWgPil6vnWb7Vo4EKb0h99LuaHUU0+oIXURwunoemxEORRiBI9c8+TH01NiIB9p/WScA70kpq4QQbwKHr7+rYQc6RzfFACGPOVLKjusOm4UQ9wE3oRkbhxvnSI7zPPA8aMso1/41F+MeK+Vl+SHLKOP3bmbooeC1+14Vk2mM24rbXIviM6LqtB9X9PYihG0cTZnaH3Bs7VBaovcRVzcUgWbJti+Z6MzjyDMZ2NbYRdUuwByv1VeZ0/sc7rdG8/uXkoO2xx89gJbMVZyWEahp0gQMdrmoNZlw46aWWi5Luoo/WcexeGw+/axmpFRZuqw/0LEWSuCj22cYw2WT3qAjhzbdD989xc7+WvbAsM2NJF5YBPFZQHCg46YLN3HN0mtYcUhL5R2TMoZ/TP8Hyw8ux+VzceOKG0Ou1YSJJDSv0urzVhNl0FynNbXLiYsdGWK9d0VpaSkvvPBCj/qGY34HzZLm5mYeffTRHz1WOBTFw8SjglNOpx69k7KyCv7+d839nZqSQnmHQK/ExBIcjhg8XiMet2YcG6sOoWu148jWsgcUZytR+7fhTuiNq3cmJ0w5ipETjsJg6r5AVDg8Hg8GQyDLYOvWrbzzzjuAJCVlN/0HHF7sa8/uMZSXd6UdcipnnnE6/3rmcaTeiDsptYt+oRw7dSrDhw7B7XSwacnnFH3yfviO2b0OO5ZUdNjzhoe05/Xvj9VpJzc3l7wJk9EbuhaC+yVRVR81JcXs+GYF6z4OOGtPueE20gb2Iioqt8MT83+WcMvF4bC37Eavi8JoTEFRFFS1ffJXKCsrIz09nV27trBixbeUlx/CZGohLk77bVRU5AKB69XpdBx//PGMHKlpxui7yMzriNfbgqLoaW7eis2Wz86d95CQMImkpGPw+VrZtfs+CvIfRKeL6vZafD4nOl3ob01KHz6fk5aWXVit/diz58Fug5QB4uLGUjj4KbzeJrZsuTbEoDEak4MC7cOh10cHeRIT4o8iJfV0eiUdi16v3cN/KWXVni6jHADmAl8CB4DLpZSfCSEGAyullIcNr+0QazFISrm7re0VoKxzzEaYfc8BbpFSjjjcOG3eillSyokdjlsNjDhczMZfn+3O2NjC0EN7AufkFSTVTKI6JTSaPHp7ESJqLE19giP/86xTGXNsAZ8t+A5XY2AiFFIghfY92Bwu7J0CN6dtKydmfBT90jbiVmHv22kcSIbstr+zvKJ1rPj2Hq4tX4yqxFCb8TSxVX/G6NziH8Oni6MuXQs+PeB+D3PlJiqOvyQocK8zR03Zhqk9ENLrhscHg117wtre30ZZqpmJa+ow39Z9IazTPjyNvY17/RoS7bQLWp2QfQKfHwgO2vrqd1+RbA02qo6EhQsXsmtXwCXfr18/TjjhBJ5++ulu9tK4+OKL6dMnVIznm2++4csvv8RqtdLaGngSu/322/nkk0/8AWHjxo5l9ZrgiTgmJoYTTzyRsrIyRo0ahcVoRFU9NNhXUFLyDwb0v5O4OM2T+eKLL1JSEiyd/e/Sv39/zjnnnJAbb319PQsWLMDVRUVLgCuvvJL9+/fz2WehDkwhfAghmTdvDjExadTWruX773fx/fcb8HgseDxHbuSYyg+gb2lGeFzINq2b1qx8dK3NpOtUZtxwJ73TQosFOu129hSt5otnHw9qN5jMpOUVMHLGTBqrq8gcVMirN1+Dz6tNaCddexP5E7sVSY7wK8PpdNLY2EhtbS1FRUWMGROIs8nP71na7n8rHVVeg9tVvN5mFMWIoph6bHD+qgJEhRB3AzcAZYAFGCCldAshLgEuaQ/G7ME4b6ItMFyKlkWyiPDZKCcCG6SUlUKIfOBd4B0p5b2HG0cI0QvYg7bE8ylwL1o8x2GzUf76bC6mPVbKwhgb4/ZuYVgHY0PnhISGyV0aG4p5FI2dMkJTU1OZM2cOT176Ou7mN0L2AxhYWs229OCnsfG7D5E6soGcdG1SL8zRJsL0Gkm9DdbM2eKfuOtSH8Jn0Ap1tdfzkCjU9HnZP17FCu3GuvaUy2hu/CDseeTm3k5Wn0u0N+v+AZ8GV6+VgDc6EcOYq2FS95VtfaqPJncT8eb4kPb3dr/HzNyZuH1uJryh/Rk9f9zzjE87siDMjng8Hh544IGgtnZPhZSSzz//nLFjx7Jz506++EJTMDWbzdx6a7c2r39sg8HAG2+8wc6dO7nllluwWCx43W5e+/P91GzbiOimbDnAuDPPZfV72vffb9Q4GqsquOChv6HodGz7ehmLnn6M1qw8VEsUwuvBUFeFqbYcVafHmZqNztmKu1foZDt5xFBWbggva95OVlYWKSkpTJ8+nR07dvD2290/XXVm5syZDB8+nMrKSp59NlisbdSoURQVFQVf67hxTJ8+HZ/Ph16vZ9myZRStWYPY+B2O9L74bIGlHLPJxLxrrsFiC9bcaKyqZO/6tdgSEhgwdmKPztPjdtFcU01CWkbY7arqY+uKrxgw9ihM1u6XUCNE+C3zqzI2AIQQZ6LpVrwjpTzU1jYLaJBSftTDMRKAF4HjgFrgVinlQiFEH2AbMFBKWSKEeAS4AC3+ohJ4DbhfSi0AoKtxOhznWOApIIuAzsaB7s4tYGxYKCsrCDE2xu7byvCDu/3vFWcriQ0nhBgb+qY6LKX7QImjOS9UNXL+/PnsWPIVn77wWMi2ZK+dgqQqVjT0Jbu6AZvTzZbMwNP9DQXaUku7sRGOjqqeMdVPItQmRivjWdJLS0k9/cADPFu8GIBrhqRyWpyHF2uM1HsFN6S4eKXWyIWJbkaP+oCYmLZS9fPDxJzcVQO6X4dbGaC1tZXly5czefJkHnnkEQAKCgqYPn06MTExQfVpfkqkqrLhs49Z/kpo5P+RMvS4E9n4ZcB7MOuRp2moKOejR/4Y0jdvwmROvOp6Fn32GUIITj75ZACqq6t5+umnMZvN3HLLLdTU1PDiiy8GxSZ0RVxcHA0NDYwbN47hw4fjcDj45z//6d8+bNgwTjst8LtQVZVVq1axdGloBeLhw4czc+bMbo+3/4f1xCb37tIgiBAhws/Pr87Y+K3T2dh4ccIM3IZgfYP0+ipO2fQtupZmpF5PYv30EGMjbr8en3M1AM0Fod/fOW++xZLh/XGHefq9JH4dUcluPvxmMHkVdVTGRLGpT8DYUFC5NHcd12bGUWQJ757uLCHemSu2nMY9tfVsiDZRP1yLf/DmPM2airWcX3A+p310Gi8e9xSj09rcysXfwksnBgaI6gX/t1VLNfgVsWDBghAhmzvuuCMo5uDn4NFzTg56H5eSSkNFOTNvvJPc0ePY/vUyFj0ViPcwR9lwttg7DxPCSfNuJn+CVg1397rv2Lh4EUOO0TwEeeOO8gtGHQlr1qwJWQrJyclh1qxZ3e7n8/moq6sjISEBXRfHffnll9m/fz+FhYWcccYZv+oKmxEiRAjwSxkbXUbOCCF6nBwvpexeJem/kTA3y9J4beIXXjeqIVRoCWBv0vdkd13wEgCzvRW31cyw4kp+6FAWPC7FCSc9yuC3tclpwNgJbCoPLN2oKDy/ZyzzLd9xcmaoG/33eefy5GEeYL+MsvKJLYo/ZwUikKd//wHTz9AEi4IKmNXtCxgaMRlw/ZFFXP+cNDQ04PP5UKRk4Yt/p7o1OKJ65syZP7mhUbzpB2KTexPbO4Xn5l5IS0MgViUhLYNZjz6NogRPxgWTplIwqetKoW6nA73ByN8uOBPV5yVn2EjOuC24IF3/0ePpP/rHLyu1M3bsWMaO1WTR3W43qqpiNh8+pkKn09GrV/eBloczWCJEiPC/TXdhup0j6YyAgYAkowJ4ABfwmzM2uvP3CKlCOLEt1UfWISdJzQ5qosOnu/oUhSS7A7vZSFqDPcjYYL6W0Zt8i56qP/+ZzL/8BS44I2SMLK+XzZOf4amGjTy3STMSrhx6JQetJ0NpLVPio8NWRIytfJBGnQ6l89VtehMmzoPeA9n29TJqDxYz6bzZLLrlHOzeQs7O2gz5J3Xzify8+Hw+HA4HNpuN1tZW/vKXv3Tb//rrrycmRsuqWf7KC/TuN4DsoSOw2KKRUtJcU401Lp6G8lKETkdieuZhz2HL8iUhgYftTLt4LsOnnxx22+EwmjW5mP9b+CH1FWXE9vplCowZjeGN5QgRIkT4OejS2JBS+nMMhRAnoelpXIcWAwEwFvgrmhz4b4c2h4bswg2seM0IokAIVCU4j9xavAuBwOryQIwBW2N/7LG7g/rs69sXj7MBvU+z2aZtPcDSQdlBfRIvmk3iRbO7PMVHt0/i2rTRXJ0ziauHX+1vb1cHLYy2hBgbs9ISWVSyDYB+psASjuljK3t1CeQ8M551Ax9n1ftaKt3aj94FOkx807rOcHY7HZRs3kju6J+uUmE7zz33HOXl5T3qa9u5gSuefZmvX3+RYy65gicuODNo+x8efJzXbrsuZL/zH/griZl9QtJDpZS8fe9tHNq+JWSfjgw7/qcxxOJTQr1VESJEiPBboKc6G48AF0spO9Yt/0YIcR3wT6DrqlS/MUriBtC3eTVSSGqTVwdtEx4XQ0sqabSYQHqxOFKRipfkqu/Z3y8OAJ9Oh1dRMLQZG/EJLvqX1zHouvAZHWm5uZTt2RPSvnX5VwybHn6SGxJt5e5+aSQa9MzboaVQzsvqTWPJRL4p+4bRbRLExcvSqS/XPACjEg5StCO8wGrjCS8Qa46ltbXVXzWyI+/+8U7Kd+/kxKtvYGA3Swbh2L9/PwcPHsRk0qR8O9ZSeOWVVw5raJjKi1E8LvQtmhDYgjlaBs7BraF1CcIZGgCv36F99sfPvZbCqccjpWTDoo9JG5AfYmjkjh6PqvqYNvtyYpO7Vw2NECFChAgaPTU2soGWMO2taBkqvyFk2//hPRv/GtmXM1p2kOwLDY4QUsXq8mA3GQE3Hsc3TCr6mlj7QQzeYezKy0NRVcqTEmkeMITHJ+zhqVQzltJckk8NX7743Ac0133VgX1BIj7fvvN6iLFhFAKLTuHkXrEoQuBVpd/YSDUZeHjKw0x4YwJ5LitENVO/JyCQVVTX9VLCe28uwXpAZceOHUyYMIHjjz/ev81ut3Po0CGETs9nTz3KZ089Snr+IAaMncCIGd1nI4AWWNiR6O1FjL74ajIH5LFvn6btZrNayYiJYkdFNQPzBlD86bugqggpmXn9rfQfPR6P08kTs87yj9PUpriY0q8/QqejvIN8d3xaBgKYOusy3nvwHn/74gVPsGHRx9SUHAg5z0nnzWbMzLNC2iNEiBAhwuHpqbGxBnhCCHG+lLIUQAiRDjwGrO52z98gy4aP55wNK0I3qCo6VaK0yYL7nGvwCi3zYNCWrezKy6M4KZbmLK3YWKojF8N5t9G7B5kdydl9ScrM8hsbjuamoO21bi9uKbkiPQmlbQlIrwj2Ti7kwHer2L5qOf3HTmD59Pf5YdsJbecbalBlRdVj0XkYGFtFstnOgt3jsCYksWOHNll/++23eL1eZsyYgVRVLc00uwDQDAWA0h1bObRzO9vr7ezerS0jhcsMsdtDszLsuYUs/W41fBf4sxLrV1KKpjd/cHuRXyvw/xZ+5M/KMJjNJGf3CykVPfkPF5M5sJCKPbv49t2FHHfZ1UQnBureTDz7D3zz9mv+9+EMjRve+p9x3EWIECHCz0JPjY1LgA+BA0KI9hrM6cBOIPwj+X8pPUnYq4+OD9suAEVKnIZA8Oj+RDOJzaBvUytsTApWxZQ6Y4/r+yg6HXOfe9W/VNBexwDggyotM+KHpuAaA67Kcj5/UqvC+tlTjzL0su10FJY7+sLLWP5KQM00P6aKwZl6+P17kD4S/YVnUrJ7F/Qb7O+zdu1aehl1fPn+O5Ac0EjwRsX4lzPs+SP8hgbgF9maMXIoSz5bxNmXzeX1Dz8OyfqRhmDD68zpx7J4e7BQFABChKR/nn3Pgzx10dmcfus9WGPiSOnX378tJXcAZ9w6P2SYcWf+nnFn/h6pqrx0/RXUl2t/3rMffZbNyxYz8Zw/hB47QoQIESIcET0yNqSUe4UQQ9BEtPLR5tVtwBL5GxPq6ONLovTw3bpEp6ocTApkolTFWlEB0fYx+aKCa3s0NTURG9t9obaORMUFDJ3Hzz+d69/8GABHWwzI/H6pFH3yPgWTphIVF89L/zeXpMF1xGbZObQ+3W9obF02gqtefAlzlI3BU4/jqYvOBiDz9lWQHAgM9bpcyCgtriNBkdS1eUM+XfVdkKEB4OgzgAzp4pDo2lOzaP1GSE7ntY8+8RsaUbt+QPF5w+qSLH78IQDS8gZStlMLcD3jtntJGxAqSWyyWn+0F0IoChc9tgCpqn4j5ugLLvlRY0WIECFChGB6LKsoNRZLKZ+QUv5NSvnlb83QgLDyGkeETpVkJAaXp9cPbmV37/DekMceeyxsSfPuyByoSZNLqeL1ePCqkgf2aYGUFV/+ixWvvciCORdQunM7ABkTK4nOaKFgZqBWSJ1uEEqbVkhHuebY5ODUy1GnnOHXFHHt3ITO3rngrmYstBNkaEgVS/HObq9F19KM0lYR01CnFXu58847OXvGCdi2r/f3O/P2e7nhrX9xxfOvkTNsJCZrVLfj/hhEGG9JhAgRIkT49+npMkq7RPgJaAGhQUn6Usr7fuLz+g9y5NZGklSxl2plvXWqZOLVN/LWvYE6G1HZLvboEtA31eGNCa1ZV1xcfFjRpI70ufpW3nnhecZ9v4IfvvgX/8gJVKzc8FZA8uSdP16P3tK1PVhXV0dKipZRceoNt+MJU4xryh8uploY2LJrD8LrwVR1iFZbsCfmqgUvs2PPXj7++OOgdtvO7xFSErV7Iz6LDXdSCqo52EiwlmjGSFxKKhc/9hQIgRCCgWPGkb/wQx47bybxaRl+PQprbFxPPqIIESJEiPArokfGhhBiHFpRMxfQCygFUtveHwB+M8aGrv/nUH4Bp2xaxYfDu64EqSL84lj9bNFsa67Hq9Pz+gmncZUpeBlB3+Yu8VpjQsYBgiSgq90e3iyv4+o+yV1KPp+2aT+MPY6Rm75hTpOgtKoBgPFFwTUqBp63B50pVBZ9fdEpALz11ltceumlNDQ0UNLsYNq0aSF99+/fz5bdezUjAFDcmlKn8HroPyCPISNGYI2JpW/fvv59oqOjGTGkEOvQgQw59gR2fLMCt8PB4ONmUFtbS5+sLDweD1JKPvrzfAZPPZ7cMeMRneqXKDpdJDgzQoQIEX4D9NSz8TDwOjAPaAKmoaXCvgH84+c5tf8sKU31DN/5Pd/nDQ+73acoKKqPxJoaMsaMYFvRCh67bD4Abx5s4WpgX2Z/ou2NbGlK1XbqUN67hA30YURbs55Gj5e3K+r5sraRlfV2piXGMMhm6fYcS9L7Upqa7X/frySwZDFszvaw+1RXZdHaqnkm6uvrefjhh/3bvv76a+bPn8/GjRvJzMwkOjo6KDX1+LnXkpzdj9dunQfA7+66yy+EFRcX8DjMmTMHW4fKncNPOMX/OqqtvT0z5Xd3/anba4wQIUKECP/99NTYGIJWSl4KIXyASUq5TwhxC7AQzRD5n0JVFFB9GJwuFn78AR2jNBxtKxfvnaTVi8h+4WZUoSK8HqRem2SHjRsHu2zU1dXhcrm4Y3cp71YGam3Uur1cta2YB/qnE2cIfE11Hq//9fszgsvX2FqaEIrKqAuT8XQ633VrZ3Lllbdis2lnet994Z1R7aXYY2JiaGoKpNeec845FBQUBPXtrLjZvm+ECBEiRIjQkZ4GiHbU5a5EK90OYAf+JzWWfUJbRHnwd5fy0jnzQrYfSsnyvx6ZeAhFKlj3b/O3zT3uRi6//HIAXC4XTV5f0P5nb9zLe5X1/LO0Jqj9gCM0rgLgpK/eIbqlmZFza/GYvg7ZbrWmExOTjKIoKIrCsGHDur2+joaGXq8PMjRGzzyLkSef3u3+ESJEiBAhQjs9NTY2AKPbXi8H/iiEmAU8AWz6Gc7rV0LXwZX66lI2pffDZdTiMzqXo3/jtMv8r//Vqi3FKF4PNSbNeDAoBn8xrB07duDr4lCiU8DqQac7bL/BBzZw/Zuf4PUFGyffbzgRn/cBrr32hqD2k08OFA4bPny4P1A0HHfeeWfQ+8nnzY6khUaIECFChB7TU2PjDqCs7fWdQDXwJBAPXP4znNevglh7U5fb3jrxD+xNTve//3rMcV329eoM1McksHjSKaxJLuKLjC+0NMu2gMiDBw+SYQ5fDl3pFCM6Z2txSJ9X5VkMuXgXbnewhPqmjcdhtydRWBga6KrX6zG1BbKecsopXHjhhSF9QFP+jBAhQoQIEf4deirqVdThdTVw4s92Rr8iBu3byoY+P9CUdBWIYGOgxWhGeAPxExsKx3c5jk9R+OLoMynvnUlM9Td8cMwtIX3KysvplFEMQHOn5ZVwtGfFrPqm0N/23bdn4/WauPPOO9Hrw3/N1157LQ6HA0VRsFqt3H777Xg8HtavX8/SpUvJy8sLkRiPECFChAgRjpQe62wACCFGAf2Af0kpW4QQUYBLSuk9zK7/lehVHybHevSeUrzG7JDtdqstdKcwtFpslPfWCp39ZfKjDErshVtVqXAFwjgPVFRii03klsG53LVHcyIlGfTUeQLGxubmgBR52dFDWdvYwucbrg053p7dY/B6TcyaNatLQwMgKiqKqKiA7oXRaMRoNHLUUUeRnJxMXl5ej64vQoQIESJE6I4eLaMIIXoLIdYAa9GyT9plJv8KPPozndtvhooOst5vVWpLM9duL2HM6u3INrEqt06PyePmdEvgK4lBZdP+A0gpqaqq4qoftFojafXV3HfvvXz++MNMIrQgnMOhSaLn5OT8qPNVFIX8/PwudT4iRIgQIUKEI6GnMRuPARVAIlpZ+XbeAY4Pu8dvin9PlX3luMBHtKyuGYAP24S4NvYdBEBxUiq10XEsWLDA33efR2WTOZrU5Ru5Y+Hb6Eo0ldLjt60FQK8PHyza0PA/mSAUIUKECBF+pfTU2DgGuENKWd+pfS+afHmPEEIkCCE+EEK0CCGKhRDn9WCfpUIIKYTQd2izd/rnE0I82bYtu61/x+139fQcw2Fq+c7/2uxydtv3ye0PHHa8L2sC9UVW98pgSYcCZF6vl9+vXcIZG5YH7fPJ0KPYnpatnYPXQ0xMJeMnvA1Ac9N0f7+vV54PEFkCiRAhQoQIvxp6amxYCNbaaKcX0P3sG8zTbeP0Bs4HnhVCDOqqsxDifMLElUgpbe3/2sZyoHlZOhLXod/9R3COIViaP+Oxv2oiWDZHa7d901xV/tcf/nBN2D4XbN4f9H5Pp+qpcQ47yc0N9K88GHZ/IVSGDlvsf3/gQMcgUu0rHThwYLfnGSFChAgRIvxS9NTYWAnM7vBeCiF0wC3AVz0ZoC2Y9EzgLimlXUq5CvgYuKCL/rHAPcDNhxn6LKAKCFWy+okQQHJ9LQCjtnUvK6KgsqzoIp7c/gBDm3cc0XFGFgf3n7Zjfdh+A/K+CXrv8ZhwOKLp2/d67rzzTk499VQKCwvD7hshQoQIESL80vTU2LgZuEwI8SVgQgsK3QZMBG7r4RgDAJ+UcleHto1AV56NPwHPosWKdMcs4JUw5e6LhRCHhBAvCSGSwu0ohLhcCFEkhCgKt70jaTVVvHTvjejsb3Tfz1VNQcs+fle1GPOdlYcbNgg1Kjr4/IC5Kz4MakuVpSQnHwhqa22NpWjdaeRkX4Ver2fEiBF+DY8IESJEiBDhP02PZiQp5TagEPgWWAyY0ZYthksp9/bwWDagsVNbIxDduWNbiu1ENOGwLhFC9AGmAC93aK5BUzvNAka2jR+2douU8nkp5Sgp5ahw2wG8zfn+19kVpazJ73rV6PUhfck64wntzenPh1Qx7cyy0cFxFfdMGhP0vr2YWXJTnb9tJu+FjCOlzq9GGiFChAgRIrl8Q68AABz/SURBVPza6LHOhpSyAm1Z48diBzrXWI8Bmjs2CCEU4BlgnpTSe5j0ywuBVVJKfxCElNIOtHsqKoUQVwPlQogYKWXXkqAdiI6uprm5F+n1zajuhKBtgtCS7e2MirFC4vFwRwUYuq/YClBgs1AxdRh/3ldOgc3CuOQ4htx8M3q9npaWFuLi4rj33nuDcmGiCb6EtWt+z3HHHcfYsWN7cmkRIkSIECHCL063j95CiD49+dfDY+0C9EKI/h3ahgJbO/WLAUYBbwkhKoB1be2HhBCTOvW9kGCvRjja5+rDikaItq7Dhn8OgMHrI6kpVK/sxJjgCf/yFe/xcmEOse3VWTsYGguH9PW/7ujJWDxqgP/1LX1TOTVZK9FutVoxGo3Ex8cjhCApKYmjqtb5+3o72Icjhi/kttseYOLEid2Kd0WIECFChAj/SQ43Qx2ge5EJ0bZdd7gDtSmOvg/cJ4S4FBgGzAQmdOraSHAl2Uw0MbGRaDVZtAMLMQFIp1MWihBiLNAA7Ear3fIEsFxK2XkJ57AIQIrQSzsnsZXPmgJOmt4lu5meFBt2DHOHpZQCm4XlY/LQC0Gu1Ry2f2cuvfRSSss+YvCBO/mQsxjFug5b/z39jwgRIkSIEOGX4HDGxugOrwWwgv9v797Dq6ru/I+/vycJBCL3W7kICKJQQClitWO9FalaOyoyUxW12rE/bB06OtbHUVs74K221daxrbaMVHnQWutUtLWtV7BavFIVKYoIFhHkfjMJCZCc7++PtRNODufkxtknIfm8nmc/nLP23mut/T0hZ2WttfeCqcCaZpZ3OfArwt0jW4BvuvvSqHfkHeDT7r6alEmhZlbzrbwh7bHoFwOPunudYRhgGGFyaV/gE+AZ4Pxm1dYhmRKil0eGzpEECS4/uC93fxRuc60sL8uaxWe7lTCxZ1euHRZWVR1Z0vDwSp0q+BbWrr2DkWzhWurewdu162ealJeIiEhLqLex4e517r00sySwxN0/aE5h7r4VODtD+mrCBNJM56wiwxCIu1+W5fiHgPpvGWkkw8FDiCo6wOaUGSffO3QAx330Ls/9Zm69eRQmjAePHFbvMfV56eUTal+XlIxg1MjbWPS3KQAUFHRsdr4iIiL5ooH+BriHdk5BEpLRiMgnu8Ocjf6FCXpt3xxj2XWHSY757B8xK+CYY56kINE5tnJFRERySQ9jqIc5JKIv/EQSqqJovbnxTQCWLdx3EbRcKi39e936RPNHDioZQadOA2MtW0REJFea09hoR7MSnQKqwZ3ClJ6NhIUXFaWNupO22V5ftHfEaeDAC2ItS0REJC71DqOY2e/TkoqB/zWzOguEuPuZua5Ya9Bn40YSPoBE1LyqToQhlZMPPhmAw4/9POtXLKewKPcP1NrxyeLa10eM/SV9+pyS8zJERETyoaE5G1vS3j8QV0Vamy7vLqKkrAzf1YOC6Dle1VHPxhcGfwGAAYeHxc56Dxma8/IXLTqn9nXXrkfkPH8REZF8aehulK/lqyKtkQFUlfDrUx/Af3R+7TBK3NInhnbs2Dc/BYuIiMRAE0QbMOHgbhQmw/BJVVq0Srr3AGDQqDE5LTOZrKh9/bljn81p3iIiIvmmW18bMLJfCQVJZw/s07PRrW8/vvaTX9C9X/+clllVVV77unPnQ3Kat4iISL6psdGQpGNrw1Lx1Rn6gXoOGJTzIvdUbQegd++JOc9bREQk3zSM0gBzp+LSK4HMjY04fLjqFwAUFqYvkisiInLgUc9GAyy5d0n5nqVw58l3xlZWMrmLtR8/zPoNjwFw8MGXxFaWiIhIvqix0YAJg7rXvi7e43TtEF9vw3vvzeDjdb+tfd+505DYyhIREckXDaM0oN+ze9ei2x1z0yy1oQFQUJBxbToREZEDihobDajasKH29Z6CfRafzak+fU6r894s3vJERETyQY2NBnj13jkbbjC+7/jYytq8+Zna1wMHTo2tHBERkXxSYyMbC0/x9D176iQXJApiK9K9uvb1wAHnx1aOiIhIPqmxkUW3oaUAeGVlbdrK3D67q470R5QXFXXPcqSIiMiBRY2NbBwSSSe5a1dt0t+Hxheu9evn1XlfWNgttrJERETySY2NLDZ0LafzniqSZWV5Ka+sbFmd9wUFnfNSroiISNzU2MiitCT0aBSPHg3ALec2L1S7dm3kufnDef4vY9m27VUWPD+a5+YPp6JiTZ3jqqrL67zXnSgiItJW5LWxYWY9zWyemZWb2Ydm1uAtF2Y238zczApT0p43s0ozK4u299LOmWhmy8xsp5ktMLMmPx2rdgZFInzpr+ltjOgxoqnZsHXrXwGort7JG29OJZkMc0A+KX27znGFhV0AGHHo9Zx80jtNLkdERKS1yvcTRH8O7Ab6AeOAP5rZYndfmulgM7uA7HWc7u73ZjinN/Ao8HXgD8BNwMPAsU2paM0Kr757NwCzT7+fPoMPa0oWAKz+aHbGdE9W1b5+e8m/s2nTkwAMHnxpk8sQEZHm27NnD2vWrKEy5YaAtqa4uJhBgwZRVFTUIuXnrbFhZiXAFGCMu5cBfzWz3wMXAddmOL4b8N/AV4GXm1DUOcBSd38kymcGsNnMRrr7snrPTFHzdI3yF14EYEiv4RQ241Hl6XMxaix95z/51KfOxN1rGxoiIpJ/a9asoUuXLgwdOrRNDmG7O1u2bGHNmjUccsghLVKHfA6jHAZUu/vylLTFwOgsx98K3AOsz7L/+2a22cwWmtlJKemjo3wBcPdyYGWmcsxsmpktMrNF6fs87efNCnL/fI3n5g9n/oJDc56viIg0XmVlJb169WqTDQ0IcwB79erVoj03+WxsHATsSEvbAXRJP9DMJgDHAT/Nktd/AcOAgcAs4A9mNryp5bj7LHef4O4T9tmXXqdmNjb69jm9WeeJiEj+tNWGRo2Wvr58NjbKgPRxiK5AaWqCmSWAu4Er3L2KDNz9VXcvdfdd7j4HWAh8qSnlNCS9Z4PC5o04NeUWVjVMRETar6FDhzJ27FjGjRvHhAnhb+BHHnmE0aNHk0gkWLRobyf8woULOeKIIzj66KNZsWIFANu3b+fUU0/d5yGRrUE+GxvLgUIzS72l40ggfXJoV2AC8LCZrQdej9LXmNnxWfJ2oKZ5sDTKF6idKzI8Qzn1SqYPoySaF6rqZOO6rYYP+zZjxtzVrDJERKRtWLBgAW+99VZtw2LMmDE8+uijnHDCCXWOu+OOO/jd737Hrbfeyj333APATTfdxPXXX9/ivRiZ5K2xEc2deBS40cxKzOw44CxgbtqhO4ABhLtVxrG3x+Io4FUz625mp5pZsZkVRnesnAA8FR03DxhjZlPMrBj4HvB2UyaHAlh610Yzh1GSyV0cdNAovnDycnp0P5YhQ77JsGFX7XPc0KGXEzp1REREglGjRnH44Yfvk15UVERFRQU7d+6kqKiIlStXsnbtWk488cQWqGXD8v3tdjnQCdgIPAR8092Xmtng6HkZgz1YX7MBm6JzN7j7bqAIuDlK3wx8Czjb3d8DcPdNhLtebgG2AccA5zW1oolk3dA0t2cjWV1JIlGMWQHjxz/IocOvpmvXIxs+UUREWoSZZd1mzZpVe9ysWbPqPbY55X7xi1/kqKOOqlNOJtdddx3Tpk3jzjvvZPr06XznO9/hpptuanKZ+ZLX52y4+1bg7AzpqwkTOzOds4q9QyQ1jYmjGyjnWWDk/tT1kN679+f0WhWVq+nYse4Kbj17HJeTvEVEpO1YuHAhAwYMYOPGjUyaNImRI0fuM3xSY9y4cbzyyisAvPDCCwwYMAB359xzz6WoqIg77riDfv365bP69VK/fRajP7X/jY3Kyo+pqFjN9u2v1kk3MyZ+YSVduozd7zJERCS33D3rNm3atNrjpk2bVu+xTTVgwAAA+vbty+TJk3nttdcaVdebb76ZG264gZkzZzJz5kwuvPBC7rqrdc0BVGMjRrt3b653f2npEgASiU75qI6IiLRS5eXllJaW1r5++umnGTNmTIPnzZkzhzPOOIMePXqwc+dOEokEiUSCnTt3xl3lJsn348rblWSy/t6Rjh36sWv3Bg499Jo81UhERFqjDRs2MHnyZACqqqqYOnUqp512GvPmzeNb3/oWmzZt4owzzmDcuHE89VS4H2Lnzp3MmTOHp59+GoCrrrqKKVOm0KFDBx566KEWu5ZM1NiIye7dW/jbG+cCYJb5TpajjnqEjRufYNDAi/JZNRERaWWGDRvG4sWL90mfPHlybSMkXefOnVmwYEHt++OPP54lS5bEVsf9ocZGTCp3rat9PX585hZmp04DGTLksnxVSUREpEVozkYeFBbs86R0ERGRdkONjZikLiGfSLTMkr4iIiKtgRobMamu3jsTuLBQPRsiItJ+qbERk2TKmigdOvRuwZqIiIi0LDU2snjqo2NqX3eesM8K9A0qK1sOwJDBmgAqIiLtmxobWWwtH1D7utP48U0+f+UHPwJg4MCpOauTiIi0XZmWmN+6dSuTJk1ixIgRTJo0iW3btgFaYr7NyNUCvYlExxzlJCIibV36EvO33XYbEydO5P3332fixIncdtttgJaYbzPMUlqG+9FKLCwsyUFtRESkPXr88ce5+OKLAbj44ot57LHHgANviXk91CsLI7WB0fTGRvfun6Vqzw4KCjrnrlIiIhKrK6+8krfeeiuneY4bN44777yzweNqlpg3My677DKmTZvGhg0b6N8/rBzev39/Nm7cCOxdYr5Tp07MnTuXq6++WkvMH4gKEtDnqqvY9OMfN+v8ZHI3HTr2zXGtRESkrcq0xHw2B9oS82psZHHVpEPh2ehNM4ZRkslKEgnd8ioiciBpTA9EXDItMd+vXz/WrVtH//79WbduHX371v0jtmaJ+Ycffpjp06czc+ZMVq1axV133cUtt9zSEpeRkeZsZFGYMLqddSZFgwbR/dxzm3x+dXUFBQVaOl5ERBqWbYn5M888kzlz5gBhOfmzzjqrznlaYv4A5yQp6tePQ599plnnJ5O7SCSKc1wrERFpi7ItMX/00Ufzla98hdmzZzN48GAeeeSR2nO0xHxb4Mn9Or26uoICNTZERKQRsi0x36tXL5577rmM5xxIS8xrGCWL/X0kSlXVDhKJDjmpi4iIyIEsr40NM+tpZvPMrNzMPjSzBh+vaWbzzczNrDB639HMZkfnl5rZm2Z2esrxQ6Pjy1K2G5pc2f3o2di27VUANm78c7PzEBERaSvyPYzyc2A30A8YB/zRzBa7+9JMB5vZBexbx0LgI+BEYDXwJeC3ZjbW3VelHNfd3atoJqf5jY1PSt8GoHLXx83OQ0REpK3IW8+GmZUAU4Ab3L3M3f8K/B64KMvx3YD/Bq5JTXf3cnef4e6r3D3p7k8A/wCOym2N92cgJTwqtrh4YG6qIiIisWqN64nkUktfXz6HUQ4Dqt19eUraYmB0luNvBe4B1teXqZn1i/JO7x350MzWmNl9ZpbxgRdmNs3MFpnZon127scwSoeiXgCMHfPzZuchIiL5UVxczJYtW1r8Czku7s6WLVsoLm65mxbyOYxyELAjLW0H0CX9QDObABwHXAEMypahmRUBDwJz3H1ZlLwZOBp4C+hFGLp5EDg1/Xx3nwXMAig6/NN1fsp2fLLvrODGqk5WANBRTxAVEWn1Bg0axJo1a9i0aVNLVyU2xcXFDBqU9es0dvlsbJQBXdPSugKlqQlmlgDuBq5w96psq9dFx80lzAGZXpPu7mVATU/FBjObDqwzs67u/kljK7tt20uNPXQfyepKABIJPdRLRKS1Kyoq4pBDDmnparRp+RxGWQ4UmtmIlLQj2Xf4oyswAXjYzNYDr0fpa8zseAALLZDZhImmU9x9Tz3l1vRY5GXN3V27N/P+ivCI2IICPWdDREQkbz0b7l5uZo8CN5rZ1wl3o5wF/FPaoTuAASnvDwZeI0wArenjugcYBZzi7hWpJ5vZMcB24H2gB3AX8Ly7pw/hxGLR65NT6lKUjyJFRERatXw/1OtyoBOwEXgI+Ka7LzWzwdHzMAZ7sL5mY28DY4O77zazIcBlhMbK+pRnaVwQHTcMeJIwPPN3YBdwfr4uMPV212xDQCIiIu2JtdXZt01lZqXAey1dj3agN2ESr8RHMY6fYpwfinP8Dnf3fW7UyDWtjbLXe+4+oaUr0daZ2SLFOV6KcfwU4/xQnOOX8dEPMdDaKCIiIhIrNTZEREQkVmps7DWrpSvQTijO8VOM46cY54fiHL+8xFgTREVERCRW6tkQERGRWKmxISIiIrFq940NM+tpZvPMrNzMPjSzqS1dp9bOzDqa2ewoXqVm9qaZnZ6yf6KZLTOznWa2IHoQW80+M7MfmNmWaPuhpTz9zMyGRufsjPI4Jd/X19qY2QgzqzSzB1LSFOMcMrPzzOzd6PfAypSlERTnHIhi8Scz22Zm683sZ2ZWGO1TjJvBzKZbWLV8l5ndn7Yvtpia2dTod3+5mT1mZj0bVWF3b9cb4UmmDxNWpf084XHpo1u6Xq15A0qAGcBQQoP1y4Qntg4lPIRnB/CvQDHwI+CVlHMvIzw8bRAwEHgH+EbK/peBHxOeNDuF8Oj5Pi19zS0c76eBF4EHoveKcW7jOwn4EDg2+nkeGG2Kc+5i/Cfg/iiOnwKWAP+hGO9XTM8BziYs33F/SnpsMQVGE37Xn0D4zvw18JtG1belA9bCH1YJYdXYw1LS5gK3tXTdDrQNeDv6wZwGvJQW4wpgZPT+JWBayv5La/4jAIcRHi/fJWX/i6n/EdrbBpwH/JbQuKtpbCjGuY3xS8ClGdIV59zF+F3gSynvfwT8UjHOSWxvpm5jI7aYArcCv07ZN5zwHdqloXq292GUw4Bqd1+ekraY0HqTRjKzfoRYLiXEbnHNPncvB1ayN6Z19lM33qOBD9y9NMv+dsXMugI3At9O26UY54iZFRBWme5jZivMbE3Uxd8JxTmX/gc4z8w6m9lA4HTCGlaKce7FGdP0vFcS/cHeUKXae2PjIEJ3U6odQOzPiW8rLCxt+yAwx92X0XBM0/fvAA6Kxgz1edR1EzDb3T9KS1eMc6cfUAT8C3A8YYHHzwDfRXHOpb8Qvqg+AdYAi4DHUIzjEGdMmx3z9t7YKAO6pqV1JYxJSQPMLEEYdtoNTI+SG4pp+v6uQJmHPjl9HhEzGwecAvwkw27FOHcqon9/6u7r3H0zYbz6SyjOORH9nngKeJTQpd8b6AH8AMU4DnHGtNkxb++NjeVAoZmNSEk7kjAcIPWIWsGzCX8ZTnH3PdGupYQY1hxXQhjXW5ppP3XjvRQYZmZdsuxvT04iTLhdbWbrgauBKWb2Bopxzrj7NsJf2pmebqg450ZP4GDgZ+6+y923APcRGnSKce7FGdP0vIcBHQnfpfVr6cktLb0BvyHckVICHIfuRmls3H4BvAIclJbeJ4rhFMJM6B9Qdyb0NwiTxQYCA6If3tSZ0K8At0fnTqYdzS5Pi2Nnwqz9mu124P+i+CrGuY31jcDrQF/CX9wvEoawFOfcxfgD4FrCSuPdgXmE4VfFuPkxLYyu+/uEHubiKC22mLJ3KOx4wnfmA+hulEZ/YD0JY4flwGpgakvXqbVvwBDCX4KVhG61mu2CaP8pwDJCF/XzwNCUcw34IbA12n5I9Nj8aP/Q6JwKwu1Zp7T09baGjZS7URTjnMe2CLg7+qW6HrgLKFaccxrjcVEstgGbgUeAvorxfsV0RvR7OHWbEXdMgamE78py4HGgZ2Pqq7VRREREJFbtfc6GiIiIxEyNDREREYmVGhsiIiISKzU2REREJFZqbIiIiEis1NgQERGRWKmxIdLGmNn9ZvZES9cjlZmdZWbvm1mVmd0fUxmt7rpFJFBjQySHoi88N7PvpqWfFKX3bqm6tbB7gd8RHgh3RUxlXAFcuD8ZmNklZlaWo/qISESNDZHcqwSuMbM+LV2RXIpW+G3Oed0Ji2895e5r3T191ciccPcd7r49jrxFZP+osSGSewuAVcAN2Q7I1NNhZkOjtAlpx5xuZn8zswoze9HMBpnZiWa22MzKzOwJM+uVoYzvmtmG6Jj7zKxTyj4zs2vMbGWU7xIzuzBDXc43s/lmVgFcluVaepjZHDPbFuX1rJmNrrkGwiOqAeZHeZ6UJZ8OZnarmX1oZrvM7AMz+4+U/SeY2atmVhld10/MrEPK/jrDKGb2vJndHeW52cw2mtnt0SqkGT8TwgJhJVE93cxmNHSN0f5uZjY3KqMyqvuVKfsvM7Pl0b5NZvaUmRWm7P+amb0T7V9uZv+ZWs+Gzhdp7dTYEMm9JGHRqW+Y2fAc5DcTuBI4hrBQ2MPA94BphNVhRxPWSUh1ImF1xomEBZm+SFiQqcbNwKXAvwOfJizm9EszOyMtn+8T1g35NGENoUzuj+p2FvBZYCfwZNS4eSmqH1E9+kdpmcwBvgpcBYyK6rcdwMwGAn8G3gQ+E+07P6pffS4AqoB/AqYT4nhulmNfivbvjOrZn7AgVUPXCCGeY4EvAyOBfwPWRnWfAPyc8DkeTli34smaQs3s/wG3Ej7TUcC3gf8CLm/M+SIHhJZeTEabtra0Eb6UnoheLyBaEZHQKHCgd6b3UdrQKG1C2jGnphwzPUobn5I2A/h7Wh22k7IiL2Euwy7CSo0lhEWWjk+r+53An9Lq8u0GrndEdNwJKWndCKtOfj163zs65qRG5HNalv23ACuAREraJdE1dU6PffT+eeDltHyeAe6tpx6XAGXNuMbfA/dlyfOc6NguWfavBi5KS7sSeKcx52vTdiBs6oYTic81wCtmdnuDR9bv7ZTXG6J/l6Sl9U0/x91TJzq+DHQAhgMdCctHP2lmqSsxFhGGf1ItaqBuowg9OS/XJLj7DjNbQugNaazPRPksqKecl909mZL2V8I1HUrdGKVKT/+YfWPVkMZc4z3A/5nZeEKD5g/u/pdo3zPAh8A/zOwp4GngUXcvjeb1HEzoVbonpcxCwuqc9Z7fxOsQaTEaRhGJibu/TrgD4wcZdtd8aVpKWrYJmHtSs43yTk9ryv/lmmP/mbD0d802mjDckqq8gbysnn1NWVK6vnxq9mfLr75y9qS9b2qsasrOpubz+DPhTpvbCT05fzSz+6J9pcB44CuEXozrgGVmNiClLt+g7mcxhmj4qYHzRQ4IamyIxOt64HjgtLT0TdG//VPSxuWw3LFmVpLy/lhgN7ASeIcw/DDE3VekbR82sZx3CL9HPleTYGZdCfMX3mlCPm9E+ZxcTzmfS5vc+Xn2XlOu7AYKMpTd4DW6+2Z3n+vulxDmlFxsZh2jfVXuPt/drwOOIAxlfdndNxDmdgzP8FmsSMk74/k5vG6RWGkYRSRG7r7CzGax77MlVgAfATPM7FrCHInvkjuFwK/M7EZgAHAb8L/uXg4QDe3cbmYGvAAcRGiQJN19VmMLcff3zexxwjDANMJckVuAT4BfNzGf3wL3mtkVhMbHIGCou88lTFK9ErjbzP4HGBZd08/cfWdjy2mEVUCxmU0iTEbd2ZhrjOL8BrCUEPtzgA/cfZeZfZkwfPUCsJXQoOoCvBuVOQP4qZltB/5E6OEaDwx09+834nyRVk89GyLxu5FwR0StaBjkPMKX5mLCnQbX57DMvxC++BYA84D5hDkkNW4gfMldHR33DOFukX80o6yvAa8RJkm+BnQmTPSsaGI+XyV8ed8FLCNM+OwG4O5rgdMJczveAn4FPERuY4a7vwT8Isp7E3tj1tA17iI0QBYDCwmNgX+O9m0Hzgaeja7rasLE0hejMu8l3L1yUXT+i4Q7jf7RmPNFDgTm3pRhVREREZGmUc+GiIiIxEqNDREREYmVGhsiIiISKzU2REREJFZqbIiIiEis1NgQERGRWKmxISIiIrFSY0NERERipcaGiIiIxOr/AznBZriochNNAAAAAElFTkSuQmCC\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"\n",
"# Common imports\n",
@@ -184,23 +175,10 @@
{
"cell_type": "code",
"execution_count": 3,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "LogisticRegression 0.864\n",
- "RandomForestClassifier 0.872\n",
- "SVC 0.888\n",
- "VotingClassifier 0.896\n",
- "LogisticRegression 0.864\n",
- "RandomForestClassifier 0.872\n",
- "SVC 0.888\n",
- "VotingClassifier 0.912\n"
- ]
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"from sklearn.datasets import make_moons\n",
@@ -256,21 +234,10 @@
{
"cell_type": "code",
"execution_count": 4,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "VotingClassifier(estimators=[('lr', LogisticRegression(random_state=42)),\n",
- " ('rf', RandomForestClassifier(random_state=42)),\n",
- " ('svc', SVC(random_state=42))])"
- ]
- },
- "execution_count": 4,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"from sklearn.datasets import make_moons\n",
@@ -295,19 +262,10 @@
{
"cell_type": "code",
"execution_count": 5,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "LogisticRegression 0.864\n",
- "RandomForestClassifier 0.896\n",
- "SVC 0.896\n",
- "VotingClassifier 0.912\n"
- ]
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"from sklearn.metrics import accuracy_score\n",
"\n",
@@ -320,22 +278,10 @@
{
"cell_type": "code",
"execution_count": 6,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "VotingClassifier(estimators=[('lr', LogisticRegression(random_state=42)),\n",
- " ('rf', RandomForestClassifier(random_state=42)),\n",
- " ('svc', SVC(probability=True, random_state=42))],\n",
- " voting='soft')"
- ]
- },
- "execution_count": 6,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"log_clf = LogisticRegression(random_state=42)\n",
"rnd_clf = RandomForestClassifier(random_state=42)\n",
@@ -350,19 +296,10 @@
{
"cell_type": "code",
"execution_count": 7,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "LogisticRegression 0.864\n",
- "RandomForestClassifier 0.896\n",
- "SVC 0.896\n",
- "VotingClassifier 0.92\n"
- ]
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"from sklearn.metrics import accuracy_score\n",
"\n",
@@ -449,82 +386,10 @@
{
"cell_type": "code",
"execution_count": 8,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(426, 30)\n",
- "(143, 30)\n",
- "Test set accuracy with Logistic Regression: 0.95\n",
- "Test set accuracy with SVM: 0.63\n",
- "Test set accuracy with Decision Trees: 0.89\n",
- "Test set accuracy Logistic Regression with scaled data: 0.96\n",
- "Test set accuracy SVM with scaled data: 0.96\n",
- "Test set accuracy with Decision Trees and scaled data: 0.89\n"
- ]
- },
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/linear_model/_logistic.py:762: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
- "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
- "\n",
- "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
- " https://scikit-learn.org/stable/modules/preprocessing.html\n",
- "Please also refer to the documentation for alternative solver options:\n",
- " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
- " extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n"
- ]
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[1. 0.73333333 0.93333333 1. 1. 0.92857143\n",
- " 1. 0.92857143 0.92857143 1. ]\n",
- "Test set accuracy with Random Forests and scaled data: 0.98\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdd3hU1dbA4d8moQsBQkBK6IGENFpCsYAUiYIggpRPAVGpAlcUQVEQQVHUC6KAVEUQBcErgoJ6QxGkEy69SwmdACEkhISU9f0xyZiQCmQymbDe55mHnDN7zllnSGbNOXuftY2IoJRSSmWkgL0DUEoplbdpolBKKZUpTRRKKaUypYlCKaVUpjRRKKWUypQmCqWUUpnSRKGUUipTmiiUwzPGnDTG3DTGRBljLhhj5hljHritTTNjzBpjTKQxJsIYs8IYU/e2NiWNMZ8ZY0KTtnUsablsBvs1xpihxph9xpgbxpgzxpglxhhfWx6vUrlNE4XKL54SkQeAekB94K3kJ4wxTYE/gJ+BikB1YDew0RhTI6lNIWA14A0EASWBZsAVIDCDfU4B/gUMBcoAtYFlQLs7Dd4Y43ynr1Eqt2iiUPmKiFwAfseSMJJ9DMwXkSkiEikiV0XkHWALMDapTS+gCtBJRA6ISKKIXBKR8SKy8vb9GGM8gFeAHiKyRkRiRSRaRBaKyEdJbdYZY15O8ZoXjDF/pVgWY8wrxpijwFFjzAxjzKe37ednY8xrST9XNMb8aIwJM8acMMYMTdEu0Bizwxhz3Rhz0Rgz6R7eRqVS0USh8hVjTGXgCeBY0nIxLGcGS9Jp/gPQJunn1sBvIhKVzV21As6IyLZ7i5ingcZAXeA7oJsxxgAYY0oDjwOLjDEFgBVYzoQqJe3/VWNM26TtTAGmiEhJoGbSsSmVIzRRqPximTEmEjgNXALeTVpfBsvv+fl0XnMeSO5/cM2gTUbutH1GPkw6w7kJbAAEeCTpuS7AZhE5BwQAbiIyTkRuichxYDbQPaltHFDLGFNWRKJEZEsOxKYUoIlC5R9Pi0gJoAXgyT8JIBxIBCqk85oKwOWkn69k0CYjd9o+I6eTfxBLhc5FQI+kVf8HLEz6uSpQ0RhzLfkBjALKJz3/EpY+kkPGmO3GmPY5EJtSgCYKlc+IyJ/APODTpOUbwGbg2XSad8XSgQ0QDLQ1xhTP5q5WA5WNMY0yaXMDKJZi+cH0Qr5t+XugizGmKpZLUj8mrT8NnBCRUikeJUTkSQAROSoiPYBywERg6R0ci1KZ0kSh8qPPgDbGmOQO7TeB3klDWUsYY0obY94HmgLvJbVZgOXD+EdjjKcxpoAxxtUYM8oY8+TtOxCRo8B04HtjTAtjTCFjTBFjTHdjzJtJzXYBzxhjihljamH51p8pEfkfEAbMAX4XkWtJT20DrhtjRhpjihpjnIwxPsaYAABjzPPGGDcRSQSSX5NwJ2+aUhnRRKHyHREJA+YDo5OW/wLaAs9g6Vc4hWUI7cNJH/iISCyWDu1DwH+B61g+nMsCWzPY1VBgKjANy4fz30AnLJ3OAJOBW8BF4Bv+uYyUle+TYvkuxTElAE9hGc11AsslszmAS1KTIGC/MSYKS8d2dxGJyeb+lMqU0YmLlFJKZUbPKJRSSmVKE4VSSqlMaaJQSimVKU0USimlMuVwhcjKli0r1apVs3cYSinlUEJCQi6LiNvdvNbhEkW1atXYsWOHvcNQSimHYow5dbev1UtPSimlMqWJQimlVKY0USillMqUJgqllFKZ0kShlFIqU5oolFJKZcpmw2ONMV8B7YFLIuKTzvMGS5XLJ4Fo4AUR2WmreO5GYqJw/XospUoVSff5Cxei+N///pnkrHz5B2jQIP25bLZvP8vly9HW5UaNKuLmlna6gJs341i37qR1uUgRZx57rHq62zxy5Ap//33Vuuzh4UqtWmXSbRscfJy4uH+qTrdqVYNChZz0mPSY9Jjuo2O6W7a8j2IelhLM8zN4/gnAI+nRGPgy6V+biIiI4fTp64SH3+Tq1Zt4e5dL95cgISGROnWmcvXqTa5di8EYQ1zcaAoUMGnabt58mmee+Wdq4qcLFeSnh2vC6m5p2o4Zs47ffjtmXV5ZsgRPfPE49EqdQ8PConnySWt1adwLFCC0TGkIG5xmm/Pn7+aDDzZYl8cXK8o7/RvBpJZp2vbo8WOqX8JLZUrjtqY7+JfTY9Jj0mPK98cUn+a5O2GzRCEi640x1TJp0hGYnzT94xZjTCljTAURyYl5iNOoUmUM16+nzM6/ARlNK/wWUBgAEeG62xd8FvMb7729KnUq2+PJP1MWW4REhtBodep1AFx5DktO/Ee/g/2ZvXpT6nYXXYBhaV5uVqdNVJxoCTyaatWss7Ppv7pV2rZxbwCpv5003NaInZdPp26nx4QeU2p6TI5+TIWAe7tYY88+ikqkmC8YOJO0Lg1jTD9jzA5jzI6wsLA0z8fHJ7JxYyjLlx/OcGfXr++/bU36l5MsbqZaWnLrECti99HAzZ0Gx90zeZ1SSuU15bFMmnj37FnCI530l2b+YMtKkVnALIBGjRqlanPhQhStWs3nwIEwcItApEbqnSwaYvlhlA9MCLSur9ugBvtDOwMwq+VGAPr3W2R5sn+MZa4yoIQxNHauTd/SI2AUHC93mRr7xwKwqehpxv/3v7DtAgD1nZ1pWKIh0irtYYxuu4YdYfvgbBQAZU0BZnnNZFar1KeVly7doHfQMlhtudveLemSV3rb/ObMLhbt3QKHLNcgazo50aNSX/q1+j5N2+5BS4n470mIvAVAQSAkcEeaU2U9Jj0mPSbHPqbTp0/zyy+/MHDgQOsxRR8sy/r4f6fZZ3bZdIa7pEtPv2TQmT0TWCci3yctHwZaZHXpqVGjRpJc6ykhIZE2bRawdu1J6/OhpUvh7vRPp5I1UVx0gRG9oNQNeCCGh1zi+WvnI6ljSm57qSQUimfLh6/Q+FQV6/PHy11m7mvH+eCljLpdlFLKPuLj4/n8888ZM2YMN27cYP369TzyyD+fccaYEBFpdDfbtucZxXJgsDFmEZYr/xF32j/RrNlzbNvmmWrdpvh4uqVIFLRO/iEC+AI5k5QY5++Dnev+aVe1ZNos3zn1Yg3ggzsJUCmlcsHWrVvp378/u3fvBqBz587UqFEji1dlny2Hx34PtADKGmPOAO9iOUNDRGYAK7EMjT2GZXhsnzvdx7Zti4BawDNAMSgYT4UCGXe7PPnkk+k/UbUkfNriTnevlFJ2FR4ezqhRo5g5cyYiQrVq1Zg6dSrt2rXL0f3Y9NKTLaS89GS5FQNOn47AvUdfGLoSeTbSnuEppVSuefXVV5kyZQrOzs4MHz6c0aNHU6xYsXTbOuqlp3vWt3AzACqvCaXvw2eYXSbKzhEppZRtxcfH4+xs+eh+5513OHHiBB988AE+Pmm6gnOMQ59R4DY11XNm0ZB0RxMopZSji4mJYeLEiSxbtoytW7dSqFChO3r9vZxROFytp5DIEMxqk/7NJUoplQ+tXr0aPz8/xo4dy65du/j9999zdf8OlyiIyfhq2ZOuGXRWK6WUA7p48SLPP/88rVu35ujRo3h5ebFu3TqeeuqpXI3D8RJFqButP5rPmgLHmXlzI7NiNkLPutCzLr/W+9Xe0SmlVI749ttv8fT0ZOHChRQpUoQJEyawa9cumjdvnuuxOGRndnDwcYKDjwMXgHX0m5T2zkmllHJkiYmJXLt2jaCgIKZNm5aj90XcKYdMFP84YO8AlFIqR0RFRbF582batGkDQM+ePalYsSKtWrWy3gpgL4536SlJhw51gEv2DkMppe7ZsmXL8PLy4qmnnuLYMUupcWMMrVu3tnuSAEdMFO6XadfOg7feetjekSil1D05deoUHTt2pFOnTpw5cwZfX19iY2PtHVYajnfpqWgcv9x0grc3sqPUG/aORiml7lhcXByfffYZY8eOJTo6mhIlSjBhwgQGDhyIk1PamfLszfESBcAeS231hs5VsmiolFJ5z9ChQ5kxYwYAXbt2ZfLkyVSsWNHOUWXM8S49JXu8mr0jUEqpu/Lqq6/i5eXFqlWrWLx4cZ5OEuCgiWJs9EoaLh5k7zCUUipLIsKCBQvo0aMHySWT6tSpw759+wgKCrJzdNnjkInivehV1p8vFI7OpKVSStnP4cOHadWqFb169WLRokWsWvXPZ1eBTKZEyGscJ9IURISQkB1QtSQPfpvOJOVKKWVHN2/eZMyYMfj5+bF27VpcXV2ZN28eTzzxhL1DuysOVz3W1DEihx0rZqXU/SM4OJgBAwbw999/A/DSSy8xceJEXF1d7RrXfVU9tlhsQditN9oppfKmTZs28ffff+Pt7c2GDRuYM2eO3ZPEvXK44bFeZx+E1j9A2GB7h6KUUiQkJHDs2DHq1KkDwMiRIylbtiwvv/zyHc8ZkVc53BmFUkrlFf/73/9o1qwZDz/8MFevXgWgcOHCDBo0KN8kCdBEoZRSdywyMpJhw4bRqFEjtm3bRuHCha19EvmRwyWK6EK3wM/N3mEope5DIsKPP/6Il5cXn332GQDDhg3j4MGDBAQE2Dk629FRT0oplU3/+te/+PzzzwEICAhg5syZ1K9f385RZc99NepJKaXspVOnTri4uDBt2jQ2b97sMEniXukZhVJKZeCvv/5i7dq1jB492rru+vXrlCxZ0o5R3Z17OaNwuOGxSilla1euXGHkyJHMnTsXgFatWtGsWTMAh0wS90oThVJKJRER5s+fz/Dhw7l8+TIFCxbkzTffvG8uMWVEE4VSSgEHDx5k4MCB/PnnnwA89thjTJ8+HU9PTztHZn8O15ldJaw0vLbG3mEopfKZSZMm8eeff+Lm5saCBQtYvXq1JokkDndG4Rb5ACw4AJNa2jsUpZSDi4iIwMXFBYAPP/yQ4sWLM2bMGMqUKWPnyPIWhzujUEqpe3Xu3Dm6detGkyZNuHXrFgBly5bls88+0ySRDk0USqn7RkJCAl988QWenp788MMPhIaGsnPnTnuHlec5XKI4VfYq/LuFvcNQSjmYkJAQGjduzNChQ4mMjKRDhw4cPHiQJk2a2Du0PM+micIYE2SMOWyMOWaMeTOd512MMSuMMbuNMfuNMX2y2ublkjegl49tAlZK5Utjx44lMDCQkJAQ3N3dWbZsGT///DNVqlSxd2gOwWaJwhjjBEwDngDqAj2MMXVva/YKcEBE/IEWwL+NMfmnNq9SKk+oUaMGxhhef/11Dhw4QMeOHe0dkkOx5ainQOCYiBwHMMYsAjoCB1K0EaCEMcYADwBXgXgbxqSUug8cP36c7du3061bNwB69uxJ48aNrZMLqTtjy0tPlYDTKZbPJK1LaSrgBZwD9gL/EpHE2zdkjOlnjNlhjNlhq2CVUo7v1q1bTJgwAW9vb3r37s2xY8cAMMZokrgHtkwUJp11t1fzawvsAioC9YCpxpg0hVREZJaINLrbglZKqfxv/fr11KtXj7fffpuYmBi6dOlyX9ZlsgVbJoozgHuK5cpYzhxS6gP8RyyOAScAvRVSKZVtly9fpk+fPjRv3pyDBw/i4eFBcHAw3377LeXKlbN3ePmCLRPFdsDDGFM9qYO6O7D8tjahQCsAY0x5oA5wPLONNjzuDm5TbRCuUsoRDRgwgHnz5lG4cGHee+899uzZQ6tWrewdVr5is85sEYk3xgwGfgecgK9EZL8xZkDS8zOA8cA8Y8xeLJeqRorIZVvFpJTKHxITEylQwPI994MPPuDmzZt89tlneHh42Dmy/MnhJi5qVLCK7Cg1AsIG2zsUpVQui46OZvz48ezatYuVK1diGTCpskMnLlJK5Xu//vorgwcP5uTJkxhj2LZtG40bN7Z3WPcFhyvhEVLjtJ5NKHUfOXPmDJ07d6Z9+/acPHkSf39/Nm3apEkiFzlcolBK3T+mT5+Ol5cX//nPfyhevDiTJk1ix44dWp8pl+mlJ6VUnnX58mWioqLo1KkTU6ZMwd3dPesXqRzncJ3Zpo4ROexYMSulsufatWscOnTIesYQGxvL2rVrCQoKsnNkju9eOrP10pNSyu5EhEWLFuHl5UWHDh24evUqAIULF9YkkQdoolBK2dWxY8cICgqiR48eXLhwAQ8PDyIiIuwdlkrB4RJF2evFYf4+e4ehlLpHsbGxjB8/Hh8fH/744w9Kly7N7Nmz2bBhA9WrV7d3eCqFbHdmG2OKi8gNWwaTHVUvl4HX1+nkRUo5uG7duvHzzz8D0KtXLz755BOtzZRHZXlGYYxpZow5ABxMWvY3xky3eWRKqXzt1VdfxdPTkzVr1vDNN99oksjDsnNGMRlLOfDlACKy2xjzqE2jUkrlK4mJiXz11VccPHiQf//73wC0aNGCffv24eTkZOfoVFaydelJRE7fVlMlwTbhZC2sRBT0vH1GVaVUXrV3714GDBjApk2bAMtlJn9/fwBNEg4iO53Zp40xzQAxxhQyxgwn6TKUPYS6hcOklvbavVIqm27cuMGIESOoX78+mzZt4sEHH2TRokX4+fnZOzR1h7JzRjEAmIJlGtMzwB/AIFsGpZRybCtWrGDw4MGEhoZijOGVV17hgw8+wMXFxd6hqbuQnURRR0SeS7nCGPMQsNE2ISmlHN2yZcsIDQ2lfv36zJw5k4CAAHuHpO5BliU8jDE7RaRBVutyi5bwUCrviY+P5+zZs1StWhWw1GhatGgRAwYMwNlZS8rlBTaZj8IY0xRoBrgZY15L8VRJLDPWKaUUW7ZsYcCAAcTGxrJ7924KFSpE2bJlGTxYpwPILzLrzC4EPIAlmZRI8bgOdLF9aEqpvCw8PJyBAwfSrFkzdu/eTUxMDCdPnrR3WMoGsnPpqaqInMqleLJUt/iDcqDJFFjdzd6hKHVfEhG+//57hg0bxqVLl3B2duaNN97gnXfeoVixYvYOT2XA1lOhRhtjPgG8gSLJK0XELmNUi90qBHvC7LFrpRTw3HPP8f333wPwyCOP8OWXX+Lt7W3nqJQtZec+ioXAIaA68B5wEthuw5iUUnlYUFAQrq6ufPXVV6xbt06TxH0gO4nCVUTmAnEi8qeIvAjoPIRK3SeCg4OZOXOmdblnz54cOXKEPn36UKCAwxWgVnchO//LcUn/njfGtDPG1Acq2zCmTB2sdAGCu9pr90rdNy5evMhzzz1HmzZt+Ne//sXff/8NgDGGMmXK2Dk6lZuy00fxvjHGBXgd+ALL8NhXbRpVJqILx4G/VplUylYSExOZNWsWb775JhERERQpUoQxY8bofNX3sSwThYj8kvRjBPAYWO/MVkrlM7t376Z///5s3boVgCeeeIKpU6dSo0YNO0em7CmzG+6cgK5Yajz9JiL7jDHtgVFAUaB+7oSolMotI0aMYOvWrVSsWJEpU6bQuXNnbqscre5DGd5HYYyZB7gD24DGwCmgKfCmiCzLrQDTxKUlPJTKMSJCdHQ0xYsXB+Dw4cPMmDGD9957j5IlS9o5OpWT7uU+iswSxT7AT0QSjTFFgMtALRG5cPeh3jtNFErljFOnTjFkyBBu3LhBcHCwnjnkc/eSKDIb9XRLRBIBRCQGOGLvJAFQLLYg7L5k7zCUclhxcXF8/PHH1K1blxUrVrB9+3aOHj1q77BUHpZZovA0xuxJeuxNsbzXGLMntwK8ndfZB6H1D/bavVIObePGjTRo0ICRI0cSHR1Nt27dOHToELVr17Z3aCoPy2zUk1euRaGUsrkhQ4YwdepUAGrUqMG0adMICgqyc1TKEWSYKPJSIUCl1L1zc3OjYMGCjBw5klGjRlG0aFF7h6QchE3vvzfGBBljDhtjjhlj3sygTQtjzC5jzH5jzJ9ZbTO60C3wc8v5YJXKZw4dOsQff/xhXR45ciR79uxh/PjxmiTUHcmyzPhdb9hyH8YRoA2Wuba3Az1E5ECKNqWATUCQiIQaY8qJSKY91TrqSanM3bx5kwkTJjBx4kRKlSrFoUOHtOSGsnmZcYwxRYEqInL4DrYdCBwTkeNJ21gEdAQOpGjzf8B/RCQUIKskoZTK3B9//MGgQYOsdZk6dOigw17VPcvy0pMx5ilgF/Bb0nI9Y8zybGy7EnA6xfKZpHUp1QZKG2PWGWNCjDG9she2Uiql8+fP0717d9q2bcvff/+Nt7c3GzZsYM6cOZQuXdre4SkHl50zirFYzg7WAYjILmNMtWy8Lr2vMbdfM3IGGgKtsJQF2WyM2SIiR1JtyJh+QD/AklqUUqk888wzbNmyhaJFizJ27FiGDRtGwYIF7R2Wyiey05kdLyIRd7HtM1hKgCSrDJxLp81vInJDRC4D6wH/2zckIrNEpNHdXl9TKj9K2b/40Ucf0b59ew4cOMCIESM0SagclZ1Esc8Y83+AkzHGwxjzBZYO6KxsBzyMMdWNMYWA7sDtl6x+Bh4xxjgbY4phqSl18A7iV+q+ExkZybBhw+jfv791XfPmzVmxYgXVqlWzX2Aq38pOohiCZb7sWOA7LOXGs5yPQkTigcHA71g+/H8Qkf3GmAHGmAFJbQ5i6fvYg6X44BwR2ZfZdquElYbX1mQjbKXyFxHhxx9/xMvLi88++4yvv/6akydP2jssdR/IcnisMaa+iPwvl+LJUqOCVWRHqREQNtjeoSiVa06cOMHgwYNZuXIlAIGBgcyYMYP69bXav8oeWxUFTDbJGHPIGDPeGKOzqCuVi0SEiRMn4u3tzcqVK3FxcWH69Ols2rRJk4TKNVkmChF5DGgBhAGzkooCvmPrwJRSlvmpjxw5ws2bN+nRoweHDh1i4MCBODk52Ts0dR+5ozuzjTG+wAigm4gUsllUmXCr8ICETdwCvXzssXulbO7y5ctcuHABHx8f6/L//vc/2rRpY+fIlCOzycRFKTbuBXQDugBXgEXAj/a6i1pLeKj8SkT45ptvGD58OG5ubuzevZtChezyfUzlQ7Yu4fE18D3wuIjcfh+EUioHHDx4kAEDBrB+/XoA/P39CQ8Pp3z58naOTKlsJAoRaZIbgSh1P4qOjuaDDz7gk08+IS4uDjc3NyZNmsRzzz2nNZpUnpFhojDG/CAiXZNmt0t5rccAIiJ+No9OqXxMRGjZsiVbt24FoH///nz44Ydam0nlOZmdUfwr6d/2uRGIUvcbYwyDBg0iOjqamTNn0rRpU3uHpFS6stOZPVFERma1LrdoZ7ZyVAkJCUyfPp24uDhee+01wHJWER8fr7WZlM3Z+oa79MbkPXE3O8sJDY+7g9tUe+1eqbuyY8cOGjduzNChQxk1ahTnzlnGhRhjNEmoPC/DRGGMGZjUP1HHGLMnxeMEltpMSqksREREMGTIEAIDAwkJCcHd3Z3FixdTsWJFe4emVLZl1kfxHbAK+BBIOd91pIhctWlUSjk4EWHJkiW8+uqrnD9/HicnJ4YNG8a7777LAw88YO/wlLojmSUKEZGTxphXbn/CGFNGk4VSmZs5cybnz5+nSZMmzJgxA3//NFOtKOUQMuzMNsb8IiLtky41CalnrBMRqZEbAaaJSzuzVR4VGxvLtWvXrDfJHT58mHXr1tG3b18KFMhOd6BStmPTEh55jSYKlRf9+eefDBgwgIoVKxIcHKw3y6k8x6ajnowxDxljiif9/LwxZpIxpsrd7Eyp/CYsLIwXXniBFi1acOjQIU6fPs3FixftHZZSOSo758NfAtHGGH8slWNPAQtsGpVSeVxiYiJz587F09OTb775hsKFC/Pee++xZ88eHnzwQXuHp1SOyk5RwHgREWNMR2CKiMw1xvS2dWBK5VUiQtu2bQkODgagdevWTJ8+HQ8PDztHppRtZOeMItIY8xbQE/jVGOME6B1C6r5ljOGRRx6hfPnyfPfdd/zxxx+aJFS+lp0SHg8C/wdsF5ENSf0TLURkfm4EeDuduEjZw6+//kpcXBxPP/00YBnhdPPmTUqVKmXnyJTKHpt2ZovIBWAh4GKMaQ/E2CtJAFS9XAZeX2ev3av7zJkzZ+jcuTPt27enb9++XL1quX2ocOHCmiTUfSM7o566AtuAZ4GuwFZjTBdbB6aUPcXHxzN58mS8vLz4z3/+Q/HixRk1ahQlS5a0d2hK5brsdGa/DQQkT31qjHEDgoGltgxMKXvZtm0b/fv3Z9euXQB06tSJKVOm4O7ubufIlLKP7CSKArfNj32F7HWC20RYiSjoWddeu1f5XGJiIn369OHAgQNUqVKFqVOn8tRTT9k7LKXsKjud2Z8AfljmzQboBuzR+ShUfiEixMbGUqRIEQDWrVvHqlWrGDNmDMWLF7dzdErlDJuX8DDGPAM8jKXe03oR+eludpYTNFGonHTs2DEGDRqEu7s7c+fOtXc4StmMTRKFMcYD+BSoCewFhovI2buOModoolA5ITY2lokTJzJhwgRiY2MpU6YMR44cwdXV1d6hKWUTthoe+xXwC9AZCAG+uJsdKJXXrFmzBj8/P959911iY2Pp3bs3hw4d0iShVAYy68wuISKzk34+bIzZmRsBKWUrCQkJ9OnThwULLKXK6tSpw4wZM2jRooV9A1Mqj8ssURQxxtTnn3koiqZcFhFNHMqhODk54ezsTJEiRXjnnXcYPnw4hQsXtndYSuV5mfVRrM3kdSIiLW0TUubqFn9QDjSZAqu72WP3ysHs3buXmJgYAgICALhy5QrXrl2jZs2ado5Mqdx1L30UGZ5RiMhjdx+S7RS7VQj2hNk7DJXH3bhxg7FjxzJ58mQ8PDzYvXs3hQoVwtXVVfsilLpD2bnhTimHsnz5coYMGUJoaCjGGFq3bk1cXByFChWyd2hKOSSb3mFtjAkyxhw2xhwzxryZSbsAY0yC1pBS9yI0NJSnn36ajh07EhoaSoMGDdi2bRtffPGF3jin1D2w2RlF0rwV04A2wBlguzFmuYgcSKfdROD37Gz3YKUL8HPXnA5XObiEhARatGjBiRMnKFGiBO+//z6DBg3C2VlPmpW6V9mpHmuS5soek7RcxRgTmI1tBwLHROS4iNwCFgEd02k3BPgRuJTOc2lEF44D/3LZaaruA8mDMZycnBg7dixdunTh4MGDDB06VJOEUjkkO5eepgNNgR5Jy5FYzhSyUgk4nWL5TNI6K2NMJaATMCOzDRlj+qhC2IYAACAASURBVBljdhhjdmRjv+o+EB4ezoABA5gwYYJ1Xc+ePVmyZAmVKlXK5JVKqTuVnUTRWEReAWIARCQcyE6voEln3e1jcT8DRopIQmYbEpFZItLobod2qfxDRFi4cCGenp7MnDmTiRMnEhERAVimKFVK5bzsnJvHJfUjCFjno0jMxuvOACkL+FcGzt3WphGwKOkPvCzwpDEmXkSWZWP76j5z5MgRBg0axOrVqwF45JFH+PLLL3FxcbFzZErlb9k5o/gc+AkoZ4z5APgLmJD5SwDYDngYY6obYwoB3YHlKRuISHURqSYi1bBMhDRIk4S6XXx8PGPHjsXX15fVq1fj6urKV199xZ9//om3t7e9w1Mq38vyjEJEFhpjQoBWWC4nPS0iB7PxunhjzGAso5mcgK9EZL8xZkDS85n2S2SkWGxB2H1JO7TvI05OTmzYsIFbt27x4osvMnHiRMqWLWvvsJS6b2Rn4qIq6a0XkVCbRJSFRgWryI5SIyBssD12r3LJxYsXiYmJoWrVqgAcPXqU8+fP8+ijj9o5MqUck63KjCf7FUu58V+B1cBxYNXd7EyprCQmJjJjxgzq1KnDSy+9ZB3+6uHhoUlCKTvJzqUn35TLxpgGQH+bRaTuW7t27WLAgAFs3boVgEKFChEVFUWJEiXsHJlS97c7LuGRVF48wAaxZEt0oVvg52av3SsbiIyM5LXXXqNhw4Zs3bqVihUrsmTJEn799VdNEkrlAdnpo3gtxWIBoAHgKiJtbRlYhvHoVKj5yq1bt/D29ubYsWMUKFCAIUOGMG7cOEqWLGnv0JTKV2xSZjyFlF/p4rH0Vfx4NztT6naFChWiZ8+erFixghkzZtCwYUN7h6SUuk2mZxRJN9p9JCJv5F5ImdMzCscWFxfH5MmTqVKlCt27dwcsZxVOTk44OTnZOTql8i+bnFEYY5yT7oVocPehKfWPjRs3MmDAAPbt24ebmxvt27fngQce0HkilMrjMrv0tA1Lf8QuY8xyYAlwI/lJEfmPjWNT+cTVq1cZOXIkc+bMAaBGjRpMnz6dBx54wM6RKaWyIzt9FGWAK0BLLPWeTNK/mihUpkSEBQsW8Prrr3P58mUKFizIyJEjGTVqFEWLFrV3eEqpbMqwj8IYcwaYxD+JIWVpThGRSbYPL62qZcrIqReWwqSW9tj9fScuLo4zZ84QExNzx68VEc6fP09cXByFCxfG1dWVggUL2iBKpVSyIkWKULly5TR/a7Ya9eQEPED2yoXnGrfIB2DBAU0UueTMmTOUKFGCatWqZauMd2JiIomJidZJg9zd3YmNjcXV1VXLgCtlYyLClStXOHPmDNWrV8+x7WaWKM6LyLgc25NySDExMdlOEhEREYSGhloTC0CJEiX0pjmlcokxBldXV8LCwnJ0u5klCv36p4CsJwS6desWp0+fJjw8HIACBQqQkJCgw12VsgNbnLlnliha5fjecsCpsldhYgt7h6GwnOaGhYVx9uxZEhISKFCgABUrVqRcuXIUKHDH1WGUUnlUhn/NInI1NwPJrsslb0AvH3uHcd9LTEzk0KFDhIaGkpCQgIuLC97e3jz44IM5niScnJyoV68ePj4+PPXUU1y7ds363P79+2nZsiW1a9fGw8OD8ePHk3KAxqpVq2jUqBFeXl54enoyfPjwHI3Nlnr06IGfnx+TJ0/OVntbDTcWEYYOHUqtWrXw8/Nj586dGbZr2bIl169ft0kcOSEoKIhSpUrRvn37DNvExsbSrVs3atWqRePGjTl58qT1uW+++QYPDw88PDz45ptvrOu7d+/O0aNHbRm6fYmIQz2ojajcc+DAgQyfO3HihOzevVuuXr0qiYmJNouhePHi1p979eol77//voiIREdHS40aNeT3338XEZEbN25IUFCQTJ06VURE9u7dKzVq1JCDBw+KiEhcXJxMmzYtR2OLi4vL0e0lO3/+vFSpUuWOXpPyfcpJv/76qwQFBUliYqJs3rxZAgMD0233yy+/yKuvvnpH246Pj8+JELMtODhYli9fLu3atcuwzbRp06R///4iIvL9999L165dRUTkypUrUr16dbly5YpcvXpVqlevLlevXhURkXXr1snLL79s+wPIpvT+boEdcrefu3f7Qns9NFHkruRfuMTERMEy2i3HH1lJ+QH45ZdfysCBA0VEZM6cOdKzZ89UbY8dOyaVK1cWEZGePXvK3Llzs9x+ZGSkvPDCC+Lj4yO+vr6ydOnSNPtdsmSJ9O7dW0REevfuLcOGDZMWLVrIq6++KlWrVpXw8HBr25o1a8qFCxfk0qVL8swzz0ijRo2kUaNG8tdff6XZ982bN637rlevnqxZs0ZERHx9faVIkSLi7+8v69evT/WaCxcuyNNPPy1+fn7i5+cnGzduTBVvZGSktGzZUurXry8+Pj6ybNkyERGJioqSJ598Uvz8/MTb21sWLVokIiIjR44ULy8v8fX1lddffz1NjP369ZPvvvvOuly7dm05d+5cmnY9evSQtWvXWpc7duwoDRo0kLp168rMmTOt64sXLy6jR4+WwMBA2bBhgyxYsEACAgLE399f+vXrZ00eAwYMkIYNG0rdunVlzJgxafZ3t9auXZtponj88cdl06ZNImL5IuDq6iqJiYny3XffSb9+/aztUr4vCQkJUq1aNZt9cbhTOZ0osnPDnbrPxcbGEhpqlwkNU0lISGD16tW89NJLgOWy0+1FBGvWrElUVBTXr19n3759vP7661lud/z48bi4uLB3714Aa6d8Zo4cOUJwcDBOTk4kJiby008/0adPH7Zu3Uq1atUoX748//d//8ewYcN4+OGHCQ0NpW3bthw8mHoW4WnTpgGwd+9eDh06xOOPP86RI0dYvnw57du3Z9euXWn2PXToUJo3b85PP/1EQkICUVFRqZ4vUqQIP/30EyVLluTy5cs0adKEDh068Ntvv1GxYkV+/fVXwDJK7erVq/z0008cOnQIY0yqy3rJzp49i7u7u3W5cuXKnD17lgoVKqRqt3HjRmbOnGld/uqrryhTpgw3b94kICCAzp074+rqyo0bN/Dx8WHcuHEcPHiQiRMnsnHjRgoWLMigQYNYuHAhvXr14oMPPqBMmTIkJCTQqlUr9uzZg5+fX6p9fvLJJyxcuDBNzI8++iiff/55uv93WUl5vM7Ozri4uHDlypUM3wewDOCoVasWu3fvzpeFLTVRqAzdunWLiIgI9u/fT2JiIjt37qRSpUq4ubnl6j0RN2/epF69epw8eZKGDRvSpk0bwHI2nFEcdxJfcHAwixYtsi6XLl06y9c8++yz1lFd3bp1Y9y4cfTp04dFixbRrVs363YPHDhgfc3169eJjIxMNVz4r7/+YsiQIQB4enpStWpVjhw5kmmZ9TVr1jB//nzA0n/j4uKS6nkRYdSoUaxfv54CBQpw9uxZLl68iK+vL8OHD2fkyJG0b9+eRx55hPj4eIoUKcLLL79Mu3bt0r12L+nclJve+3v16tVUx/b555/z008/AXD69GmOHj2Kq6srTk5OdO7cGYDVq1cTEhJCQIBlipubN29Srlw5AH744QdmzZpFfHw858+f58CBA2kSxRtvvMEbb+RszdKMjjer96FcuXKcO3cuXyYKHZqi0rVhwwbq16/PtWvXSExMpEyZMvj4+FCuXLlcv3GuaNGi7Nq1i1OnTnHr1i3rt3Bvb2927NiRqu3x48d54IEHKFGiBN7e3oSEhGS5/YwSTsp1t9+ZXrx4cevPTZs25dixY4SFhbFs2TKeeeYZwNLhv3nzZnbt2sWuXbs4e/ZsmntK0vvwuVcLFy4kLCyMkJAQdu3aRfny5YmJiaF27dqEhITg6+vLW2+9xbhx43B2dmbbtm107tyZZcuWERQUlGZ7lStX5vTp09blM2fOULFixTTtnJ2dSUxMBGDdunUEBwezefNmdu/eTf369a3vYZEiRaxJVkTo3bu39T06fPgwY8eO5cSJE3z66aesXr2aPXv20K5du3SrA3zyySfUq1cvzWPo0KF3/f6lPN74+HgiIiIoU6ZMlu9DTExMvi1N43CJouFxd3Cbau8w8rWbN2/SpUsXDhw4gLOzM7Vr16ZGjRp2L7/h4uLC559/zqeffkpcXBzPPfccf/31F8HBwda4hw4dyogRIwDLt80JEyZw5MgRwPLBPWlS2sozjz/+OFOn/vM7lXzpqXz58hw8eNB6aSkjxhg6derEa6+9hpeXF66uruluN73LSI8++qj10smRI0cIDQ2lTp06mb4PrVq14ssvvwQsl+NuH2UUERFBuXLlKFiwIGvXruXUqVMAnDt3jmLFivH8888zfPhwdu7cSVRUFBERETz55JN89tln6cbYoUMH5s+fj4iwZcsWXFxc0lx2AqhTpw7Hjx+3xlC6dGmKFSvGoUOH2LJlS4bHsnTpUi5dugRYzkpOnTrF9evXKV68OC4uLly8eJFVq1al+/o33njDmmRSPu72slPy8SaPaFq6dCktW7bEGEPbtm35448/CA8PJzw8nD/++IO2bf+Zv+3IkSN4e3vf9X7ztLvt3LDXo6Gzu0jZL+60b0dlITExMVVH3LfffitjxoyR/fv32zEqi9tH87Rv317mz58vIiJ79uyR5s2bS+3ataVmzZoyduzYVCOwVqxYIQ0aNBBPT0/x8vKS4cOHp9l+ZGSk9OrVS7y9vcXPz09+/PFHEbF0YNeoUUOaN28ur7zySqrO7CVLlqTaxvbt2wWQefPmWdeFhYVJ165dxdfXV7y8vKwjaVK6efOm9O7dO01n9okTJ8Tb2zvd9+PChQvSoUMH8fHxEX9/f2vHa/L7FBYWJk2aNJGGDRvKSy+9JJ6ennLixAn57bffxNfXV/z9/aVRo0ayfft2OXfunAQEBIivr6/4+Pikij9ZYmKiDBo0SGrUqCE+Pj6yffv2dOMaN26czJ49W0REYmJiJCgoSHx9faVLly7SvHlza0f37f+fixYtEn9/f/H19ZUGDRrI5s2bre+zp6enPPnkk9KpUyf5+uuv093vnXj44YelbNmyUqRIEalUqZL89ttvIiIyevRo+fnnn0XE8n/SpUsXqVmzpgQEBMjff/9tff3cuXOlZs2aUrNmTfnqq6+s6y9cuCABAQH3HF9OyenO7CynQs1rGhWsIjtKjYCwwfYOJd84cOAAAwYMoE2bNowePTrVcwcPHsTLy8tOkSlHcv78eXr16sV///tfe4eS6yZPnkzJkiWtAy3sLb2/23spCuhwl55UzomOjmbUqFH4+/uzYcMG5syZQ2xsrL3DUg6qQoUK9O3bN0/fcGcrpUqVonfv3vYOw2YcLlGE1DitZxM5YNWqVfj4+PDhhx8SHx9P//792bVrF4ULF7Z3aMqBde3aNdMRW/lVnz59rBWT86P8e2QqXTdu3OCFF15g6dKlAPj5+TFjxgyaNm1q58iUUnmVw51RqHtTrFgxrl69SvHixfn0008JCQnRJKGUypSeUdwHduzYQalSpahVqxbGGObMmYOTkxNVqlSxd2hKKQegZxT5WEREBEOGDCEwMJABAwZYb+6qXr26JgmlVLZposiHRITFixfj6enJ1KlTKVCgAA0aNCA+Pt7eod0VLTNu3zLjhw4domnTphQuXJhPP/00w3YiWmY837rbGzDs9Sj7YHGRb/bexS0o94djx45J27ZtrZVZmzZtKrt3777r7WVWZjy3aJnx7LFVmfGLFy/Ktm3bZNSoUfLJJ59k2E7LjGuZ8Tzz0DuzM3b9+nUpVaqUAFKqVCmZOXOmJCQk3NM2U/7CEYxNHlnRMuP2LTOe7N133800UWiZcS0zfleMMUHAFMAJmCMiH932/HPAyKTFKGCgiOy2ZUz5WYkSJRg2bBjHjh3j008/tVbhzC+0zLhFbpcZzy4tM65lxu+YMcYJmAa0Ac4A240xy0XkQIpmJ4DmIhJujHkCmAU0tlVM+U1YWBhvvPEGrVq1omfPngCMHj3aZtVdpZV9yr1omfHUcrvMeHZpmXEtM343AoFjInJcRG4Bi4COKRuIyCYRSf76tgWonNVGw0pEQc+6OR6sI0lMTGTOnDnUqVOHb775hrfffpu4uDjgzj4gHYWWGb8zOV1mPLu0zLiWGb8blYDTKZbPJK3LyEtAurWEjTH9jDE7jDE7Qt3CYVLLHAzTsezbt49HH32Uvn37Eh4eTuvWrVm9erXdS4DnBi0zbpHbZcazS8uMa5nxO+8lh2ex9EskL/cEvsig7WPAQcA1y+3ep3NmR0dHy4gRI8TZ2VkAKV++vHz33XepSmrbQl4b9SSiZcZzu8z4+fPnpVKlSlKiRAlxcXGRSpUqSURERJp2WmZcy4zfMWNMU2CsiLRNWn4rKTF9eFs7P+An4AkROZLldusYkcOOVRo9J8TGxlKvXj0OHz7MwIED+eCDDyhVqpTN96tlxlV2aZnx/Ftm3JajnrYDHsaY6sBZoDvwfykbGGOqAP8BemYnSdxvzpw5Q7FixShTpgyFCxdm3rx5ADRurP39Ku9JWWb8fqsgW6pUKeuAkvzIZn0UIhIPDAZ+x3JZ6QcR2W+MGWCMGZDUbAzgCkw3xuwyxuzIYHP3lfj4eCZPnoyXl1eqER2NGzfWJKHyNC0znj/Z9MhEZCWw8rZ1M1L8/DLwsi1jcDRbt26lf//+7N5tuZ0kIiKC+Pj4fP1LqJTK2xyu1pPXmfLQarG9w8hx165dY9CgQTRt2pTdu3dTtWpVVqxYwdKlSzVJKKXsyuE+gYrdKgR7wuwdRo4KDw+nbt26XLhwAWdnZ15//XVGjx6daqy+UkrZi8MlivyodOnSPPHEExw5coQvv/wSX19fe4eklFJWDnfpKT+IjY1l3Lhx/Pnnn9Z1U6dOZf369Zok7tLy5cv56KOPsm6Yz82bNw83Nzfq1auHp6dnmhLls2bNwtPTE09PTwIDA/nrr7+sz8XFxfHmm2/i4eGBj48PgYGBGd7oZk+vvvoq69evt3cYGXr77bdxd3fPsuz7hx9+SK1atahTpw6///67dX3y3fO1atVi6NCh1rv3p06dytdff23T2DN0tzdg2OtRrGpBkV0X7+jmk7xk9erVUrt2bQHEy8sr18ss36k0N+6U/SL1IyPf7E3dbthq2wZ6BxITE++5qu69sGWF0a+//lpeeeUVERG5fPmyuLq6SmhoqIj8c/NhWFiYiIiEhISIu7u7nD9/XkQsVWR79eolMTExImK5iWzx4sU5Gt+9/r5fuXJFGjdufEevye2Krps3b5Zz585lWvZ9//794ufnJzExMXL8+HGpUaOG9b0JCAiQTZs2SWJiogQFBcnKlStFxFJGv169etmKIadvuHO4M4rownHg73hVUS9dukTPnj1p1aoVR44cwdPTk+nTp1tr3qj0nTx5Ek9PT15++WV8fHx47rnnCA4O5qGHHsLDw4Nt27YBlm/SgwcPBuDixYt06tQJf39//P392bRpEydPnsTLy4tBgwbRoEEDTp8+zRtvvIGPjw++vr4sXpz+AIlt27bRrFkz6tevT7NmzTh8+DBgGaq8f/9+a7sWLVoQEhLCjRs3ePHFFwkICKB+/fr8/PPP1vieffZZnnrqKR5//HGioqJo1aoVDRo0wNfX19oOLNVsPT09adOmDT169LBOFvT3338TFBREw4YNeeSRRzh06FCm752rqyu1atXi/PnzAEycOJFPPvmEsmXLAtCgQQN69+7NtGnTiI6OZvbs2XzxxRcULlwYsJQw6dq1a5rtbt++nWbNmuHv709gYCCRkZGp3n+A9u3bs27dOsAyodKYMWNo3LgxEyZMSLXNdevW8dRTTwHwxx9/0LRpUxo0aMCzzz6bpiouWEpqpKxHNW7cOAICAvDx8aFfv37Wb98tWrRg1KhRNG/enClTphASEkLz5s1p2LAhbdu2tb4ns2fPJiAgAH9/fzp37kx0dHSm72l2NGnShAoVKmTa5ueff6Z79+4ULlyY6tWrU6tWLbZt28b58+e5fv06TZs2xRhDr169WLZsGWCZ775atWrW3/lcdbcZxl4PRyvhkZCQIDNnzrTOE1GkSBF5//33JTY21t6hZYu9zyhOnDghTk5OsmfPHklISJAGDRpInz59JDExUZYtWyYdO3YUkdTfpLt27SqTJ08WEcs32GvXrsmJEyfEGGMtD7F06VJp3bq1xMfHy4ULF8Td3V3OnTuXZv8RERHWb6T//e9/5ZlnnhERkUmTJlnnSDh37px4eHiIiMhbb70lCxYsEBGR8PBw8fDwkKioKPn666+lUqVKcuXKFRGxfMtNLoMRFhYmNWvWlMTERNm+fbv4+/tLdHS0XL9+XWrVqmWdA6Jly5Zy5MgRERHZsmWLPPbYY2niTfk+nDp1Svz9/eXmzZsiIlK6dGm5du1aqvbLli2TTp06ye7du7P1bTU2NlaqV68u27ZtS/X+pNyviEi7du2sJTsA65lJXFycuLu7S1RUlIhY5pxYsGCBhIWFySOPPGJd/9FHH8l7772XZv+9evWS5cuXW5eT308Rkeeff976XPPmza3zlty6dUuaNm0qly5dEhFLyZA+ffqIiOWsK9nbb78tn3/+eZp9rlmzRvz9/dM8mjZtmul7ldkZxSuvvGL9PRERefHFF2XJkiWyfft2adWqlXX9+vXrU82d8f7778unn36a6X5FHGw+CmW5D+Ltt9/m2rVrtG3blmnTplGzZk17h+VQqlevbu278fb2plWrVhhj8PX1TTVNZbL0ynCHh4dTtWpVmjRpAljKe/fo0QMnJyfKly9P8+bN2b59Ox06dEi1rYiICHr37s3Ro0cxxlir9Hbt2pU2bdrw3nvv8cMPP/Dss88Clm/Fy5cvt54FxMTEEBoaCkCbNm0oU6YMkHEp8L/++ouOHTtaq5Amf9uOiopi06ZN1v2Apa8rPYsXL2bt2rUcPnyY2bNnU6RIkQzfW5GMS7Wn5/Dhw1SoUMFaFjw7N9elLCvu7OxMUFAQK1asoEuXLvz66698/PHH/Pnnnxw4cICHHnoIgFu3btG0adM02zp//jxubm7W5bVr1/Lxxx8THR3N1atX8fb2tr5nyeXeDx8+zL59+6zl6RMSEqzf+Pft28c777zDtWvXiIqKSlXkL9ljjz12T8US0yPplE7KbinzrM4kbUEThQ3cuHEDZ2dnChcuTOnSpZkxYwYJCQk8++yzjl8GPGxw1m0AevlYHjkg+VIIWCaISV4uUKDAHc0DnnK4cXp/kGCZSGj27NkArFy5ktGjR/PYY4/x008/cfLkSVq0aAFApUqVcHV1Zc+ePSxevNg6YY+I8OOPP6apALt169ZU+09ZCrxgwYJUq1aNmJiYDONKTEykVKlS2frA6tatG1OnTmXz5s20a9eOJ554ggcffJC6desSEhJCy5b/VF/euXMndevWpVatWoSGhqaZL+N2GSWWlCXGIXVZ9pRlxZPjmzZtGmXKlCEgIIASJUogIrRp04bvv/8+02MrWrSoddsxMTEMGjSIHTt24O7uztixY1PtN/n9FhG8vb3ZvHlzmu298MILLFu2DH9/f+bNm2e9XJbS2rVrGTZsWJr1xYoVY9OmTZnGm5GMSpZXrlyZM2fOpFmfzF6lzB2ujyKvW758OXXr1uXjjz+2ruvcuTNdu3Z1/CThILIqww2W8t6LFy8mISGBsLAw1q9fT2BgIK+88oq1VHXFihWJiIigUiVLdfzkWlvJunfvzscff0xERIT1jKdt27Z88cUX1g/8//3vf+nGmFEp8IcffpgVK1YQExNDVFSUdTa6kiVLUr16dZYsWQJYPvyS797PSNOmTenZsydTpkwBYMSIEYwcOZIrV64AlrLn8+bNY9CgQRQrVoyXXnqJoUOHcuvWLcDy7f3bb79NtU1PT0/OnTvH9u3bAYiMjCQ+Pp5q1aqxa9cuEhMTOX36dKbX0Vu0aMHOnTuZPXu29Vt/kyZN2LhxI8eOHQMgOjraWh4+JS8vL2ub5KRQtmxZoqKiWLp0abr7q1OnDmFhYdZEERcXZ+1fioyMpEKFCsTFxaU7Ux78c0Zx++NukwRYSpkvWrSI2NhYTpw4wdGjRwkMDKRChQqUKFGCLVu2ICLMnz+fjh3/mcbnyJEj+PjkzBewO+FwiaJYbEHYfcneYaQRGhrK008/TceOHQkNDeX3339P9Q1L5Z4pU6awdu1afH19adiwYapO52SdOnXCz88Pf39/WrZsyccff8yDDz6Ypt2IESN46623eOihh0hISEj1XJcuXVi0aFGqztnRo0cTFxeHn58fPj4+jB49Ot0Yn3vuOXbs2EGjRo1YuHAhnp6eAAQEBNChQwf8/f155plnaNSokXUGu4ULFzJ37lz8/f3x9vZO1QGekZEjR/L1118TGRlJhw4dePHFF2nWrBmenp707duXb7/91noZ5v3338fNzY26devi4+PD008/neoyD0ChQoVYvHgxQ4YMwd/fnzZt2hATE8NDDz1kvUQ4fPhwGjRokGFMTk5OtG/fnlWrVlln1HNzc2PevHn06NEDPz8/mjRpku4llnbt2lm/9ZcqVYq+ffvi6+vL008/bb0cdrtChQqxdOlSRo4cib+/P/Xq1bN+yI8fP57GjRvTpk0b6//BvRoxYgSVK1cmOjqaypUrM3bsWMDyJXLMmDGA5RJq165dqVu3LkFBQUybNs161vXll1/y8ssvU6tWLWrWrMkTTzxh3fbGjRtp3bp1jsR5R+62c8Nej4bO7pl3ouayW7duySeffCLFihUTQEqUKCFTpkzJ88NesysvzEdxv4mMjBQRy3DIhg0bSkhIiJ0jylseeughCQ8Pt3cYuW7nzp3y/PPPZ6utdmbnIZcvX7ZO+g6WeZQnT55svVSh1N3o168fBw4cICYmht69e2f67fx+9O9//5vQ0NBcmY8lL7l8+TLjx4+3y741UdwDV1dXypYtxLKXTwAAFFVJREFUS/Xq1Zk6dSpPPvmkvUNS+cB3331n7xDytPu11H7yqC17cLhEEV3oFvi5Zd3QBkSEhQsXEhgYSO3atTHG8O233+Li4kKxYsXsEpNSStmaw3VmH6x8EVZ3y/X9Hj58mNatW9OzZ08GDRpkHdVSoUIFTRJKqXzN4RJFbouJieHdd9/Fz8+PNWvW4OrqyvPPP2/vsJRSKtc43KWn3BQcHMzAgQOt47ZffPFFPv74Y1xdXe0cmVJK5R49o8jAxYsXad++PceOHaNu3bqsX7+euXPnapJQDuPkyZMULVqUevXqUbduXXr16mUtQQKWMiaBgYHWsuOzZs1K9fr58+fj4+ODt7c3devWtZYlyUuWLVvGuHHj7B1GhpYsWYK3tzcFChRgx44dGbb77bffqFOnDrVq1UpVLv/q1au0adMGDw8P2rRpQ3h4OAB79+7lhRdesHX4/7jbcbX2etiyKGBCQoIkJiZalydOnCgffvihwxTws4Xbx2PD2FSPjMycuSNVu759l2fY1t7sec+LLUuenzhxQry9vUXEcoyPPfaYfPvttyIicv78eXF3d7feoxEWFiYNGjSQX375RUREVq5cKfXr15ez/9/euUdHUeV5/PMDAmEAIUeUEwkR8DEJkYTwirA8wshGBzCiooAuig8QZICVVR46Z/Aw6wTlcXaQgwgsBxxGoqDgY9dRWOUhigYhKAbkKU4gGEhmSDQIhPz2j6qudJLuThGSTie5n3PqdFXXrXt/9evq+tV91PeeOKGqqufOndNly5ZVq33VIf/dp08fRzY9WGVeDllZWXrgwAEdOHCgZmRk+ExTXFysnTt31iNHjuj58+c1Pj5ev/32W1VVfeaZZzQtLU1VVdPS0nT69OnOcbfddpseP37cb7nloSHJjNcUmZmZ9O3bt4xkwfTp05k5cyZNmzatRcsaNm5lxv3JgV+6dImnn36arl27Eh8fz8svvwxAx44dmTNnDv369WPdunWsXbuWrl27cssttzBjxgyftviTBp8xYwZLlixx0j3//PMsWLAAgHnz5tGrVy/i4+OZPXu2c07lJc8nTpxIz549iYuLc9KBpTcVExNDv379mDJlivMmsz85c380btyY3r17c+LECcDStBo7dqzzjkbbtm156aWXnKfZtLQ05s+f7+gMhYeHM27cuAr5+pN095aZmD9/vvN2srf89wsvvEDHjh0dBYOioiI6dOjAxYsXXUmqHzx4kGbNmjmy6e+99x5JSUkkJiYyePBgfvzxR+f3GD9+PCkpKTz00EOcPn2ae++9l169etGrVy927NgB+L+GroTY2NgKul/l+fLLL7nxxhvp3LkzTZs2ZdSoUc7v+c477/Dwww8D8PDDDzuS42AJRqanp1+xja6oaoSprSU6IqJaJ8EpKCjQp556Shs1aqSAduvWrUytoqFT2zUKtzLj/uTAlyxZovfcc4+zzyNLff311+uLL76oqqonTpzQDh06aG5url68eFEHDRqkGzZsqGCLP2nw3bt364ABA5x0sbGxevz4cf3www913LhxTq1h6NChunXr1gqS5952FRcX68CBA3Xv3r167tw5jYqK0qNHj6qq6qhRoxzJaX9y5uV956lRnDt3TpOTk3Xv3r2qqnr33Xfrxo0by6T/5z//qREREarqW5LcF/4k3T3lqqrOmzdPZ8+erapl5b9VVVNTU/Xjjz9WVUv++7HHHlNVd5LqK1eu1GnTpjnb+fn5zn93+fLlzr7Zs2dr9+7dtaioSFVVR48erdu3b1dVS4o9JiZGVf1fQ94UFBT4lBxPSEhwagG+CFSjWLdunXPeqqqvvfaaI9neunXrMmnbtGnjrH/66ac6bNgwn3k2+DezrylsCX/JgoW/qTxxAFSVjRs3MmXKFLKzs2nUqBFTp05lzpw5RrwvxHAjM+5PDnzz5s1MmDCBJk2sS90j8w2lMtQZGRkkJyc7ukYPPvgg27ZtY/jw4WXsUPUtDZ6YmEhubi4nT57k9OnTREREEB0dzaJFi/joo49ITEwErBrJoUOHiI6OLiN5DvDmm2+ybNkyiouLycnJISsri5KSEjp37kynTp0AGD16tNOP4E/OPDY2tozNR44coVu3bhw6dIgRI0YQHx/vnIuv6/xyr31/ku6B8Pjds/7GG28waNAg0tPTefLJJ11LqpeXHM/OzmbkyJHk5ORw4cIFx29gifB5VFc3b95MVlaWs6+goIDCwkK/15A3rVq1CprkeGVce+21nDx5slpt8UedCxTVwZkzZ3jkkUd4//33AejZsyevvvqqkUpwgersyhMB48f3YPz4HtVSphuZcX9y4P5uiFBWhtoXX3zxBU888QRgzaSWn5/vUxocLIHA9evXc+rUKUaNGuXkO2vWLCcPD99//30ZyfFjx44xf/58MjIyiIiIYOzYsQElxz15+5IzL88NN9xAZmYmOTk5JCcn8+6775KamkpcXBy7du0qM//GV199RZcuXQArIJeXJHdLIMlxKCv3npqayqxZs8jPz3fK+/nnn11Jqjdv3pyzZ88625MnT2batGmkpqayZcsWp7mrfJklJSV8/vnnFeS6J0+e7PMa8qawsJD+/fv7tOf11193/Hc5+JMcB2uWwZycHCIjI8nJyeHaa0tn9wym5HiD7KNo1aoVhw8f5qqrrmLx4sXs3LnTBIk6jj858JSUFJYuXeoElPz8/ArHJiUlsXXrVs6cOcOlS5dYu3YtAwcOJCkpyZGUTk1N9SsNDpbkeHp6OuvXr2fEiBGAJTm+cuVKZ0rPEydOkJtbUfm4oKCAFi1a0Lp1a3788Uc++OADwJL0Pnr0qFNr8p6u1a2cuYfIyEjmzp1LWloaAJMmTWLVqlXOzTgvL48ZM2Ywffp0AGbNmsX06dM5deoUYD3RL1q0qEK+viTd27VrR25uLnl5eZw/f955IPNFy5Yt6d27N1OnTmXYsGE0btzYtaS6t+Q4lL0GVq9e7bfMlJQUFi9e7Gx7fBBIUt6Dp0bha6lKkABLMfjQoUMcO3aMCxcukJ6e7gTw1NRU51xWr15da5LjdS5QHG+bDwuSL/u4HTt2ODr8zZo1Iz09nQMHDjBp0iQzb3U9wJ8c+OOPP050dLQjKe5LRykyMpK0tDQGDRpEQkIC3bt3L/OH9OBPGhysJ/DCwkLat2/vyHanpKTwwAMP0KdPH7p27cqIESMoLCyskG9CQgKJiYnExcXx6KOPOrO8NW/enCVLlnDHHXfQr18/2rVr50iOu5Uz92b48OEUFRWxfft2IiMjWbNmDePGjSMmJoa+ffvy6KOPOrPDDRkyhEmTJjF48GDi4uLo0aOHz0mifEm6h4WFOXNkDxs2rFL57pEjR7JmzZoyTVJuJNUHDBjAnj17nGD5/PPPc99999G/f3+ng9sXixYtYteuXcTHx9OlSxeWLl0KBJaUryobNmwgKirKmUTKM4PeyZMnHW24Jk2asHjxYm6//XZiY2O5//77iYuLA2DmzJls2rSJm266iU2bNjFz5kwn708++YShQ4dWi52VUtXOjdpaLnd47JkzZ/Txxx9XoEyHkcEdRma8dvFIjpeUlOjEiRN14cKFtWxRaDFlyhTdtGlTbZsRdH755RdNSkryO9zXDI91iaqyevVqYmJiWLFiBWFhYVx33XUB230NhlBj+fLldOvWjbi4OM6ePVuhv6Oh8+yzz1JUVFTbZgSdH374gblz5zqDNGoaqWs3Tvm1qH4X2OYDBw4wYcIEtm7dClhjt1955ZVqm8GqIbF///4KI2kMBkNo4+t/KyJfqWrPquRX70Y9ZWdnk5CQwIULF2jbti0LFixgzJgxZsjrFaABRg4ZDIbQoiYe/utdoIiKimLMmDE0atSIuXPnlhk3b7h8wsPDycvL4+qrrzbBwmAIcVSVvLw8wsPDqzXfOt/0lJOTw1NPPcWECROccc8lJSU0alRvu1+CysWLF8nOzq4wFt5gMIQm4eHhREVFERYWVub7BtX01ONoB7hmMZdOTeSVV17hueeeo6CggMOHD5ORkYGImCBRjYSFhZV5w9VgMDQ8avSOKiJ3iMh3InJYRGb62C8issje/7WIuHrrbXfx37n11luZPHkyBQUF3Hnnnbz11lumacRgMBhqgBprehKRxsBB4F+BbCADGK2qWV5phgCTgSFAEvBnVQ04c3q7Rq30jP5MCUpUVBQvv/wyd911lwkSBoPBEIAraXqqyRpFb+Cwqh5V1QtAOlD+dde7gNfs90F2Am1EJDJQpvlahCBMmzaN/fv3M3z4cBMkDAaDoQapyT6K9sDfvbazsWoNlaVpD+R4JxKR8cB4e/M8sG/hwoUsXLiwWg2ug7QFztS2ESGC8UUpxhelGF+UElhBMgA1GSh8PeaXb+dykwZVXQYsAxCRXVWtPtU3jC9KMb4oxfiiFOOLUkTE/1yslVCTTU/ZQAev7SigvHi6mzQGg8FgqEVqMlBkADeJSCcRaQqMAt4tl+Zd4CF79NOtwFlVzSmfkcFgMBhqjxprelLVYhH5HfAh0BhYqarfisgEe/9S4H+xRjwdBoqAR1xkvayGTK6LGF+UYnxRivFFKcYXpVTZF3XuzWyDwWAwBBfzCrPBYDAYAmIChcFgMBgCErKBoqbkP+oiLnzxoO2Dr0XkMxFJqA07g0FlvvBK10tELonIiGDaF0zc+EJEkkUkU0S+FZGtwbYxWLj4j7QWkfdEZK/tCzf9oXUOEVkpIrkiss/P/qrdN6s6NV5NLlid30eAzkBTYC/QpVyaIcAHWO9i3Ap8Udt216Iv+gIR9vpvG7IvvNJ9jDVYYkRt212L10UbIAuItrevrW27a9EXzwIv2uvXAPlA09q2vQZ8MQDoDuzzs79K981QrVHUiPxHHaVSX6jqZ6r6D3tzJ9b7KPURN9cFWPphbwG5wTQuyLjxxQPA26r6A4Cq1ld/uPGFAq3E0vtpiRUoioNrZs2jqtuwzs0fVbpvhmqg8Cftcblp6gOXe56PYT0x1Ecq9YWItAfuBpYG0a7awM11cTMQISJbROQrEXkoaNYFFze+WAzEYr3Q+w0wVVVLgmNeSFGl+2aozkdRbfIf9QDX5ykig7ACRb8ataj2cOOL/wJmqOqlei4W6cYXTYAewG1Ac+BzEdmpqgdr2rgg48YXtwOZwG+AG4BNIrJdVQtq2rgQo0r3zVANFEb+oxRX5yki8cAK4Leqmhck24KNG1/0BNLtINEWGCIixaq6MTgmBg23/5Ezqvoz8LOIbAMSsOT/6xNufPEIMFethvrDInIMiAG+DI6JIUOV7puh2vRk5D9KqdQXIhINvA2MqYdPi95U6gtV7aSqHVW1I7AeeLIeBglw9x95B+gvIk1E5FdY6s37g2xnMHDjix+walaISDssJdWjQbUyNKjSfTMkaxRac/IfdQ6XvvgDcDWwxH6SLtZ6qJjp0hcNAje+UNX9IvI34GugBFihqj6HTdZlXF4XfwRWicg3WM0vM1S13smPi8haIBloKyLZwGwgDK7svmkkPAwGg8EQkFBtejIYDAZDiGAChcFgMBgCYgKFwWAwGAJiAoXBYDAYAmIChcFgMBgCYgKFISSxlV8zvZaOAdL+VA3lrRKRY3ZZu0WkTxXyWCEiXez1Z8vt++xKbbTz8fhln62G2qaS9N1EZEh1lG1ouJjhsYaQRER+UtWW1Z02QB6rgPdVdb2IpADzVTX+CvK7Ypsqy1dEVgMHVfWFAOnHAj1V9XfVbYuh4WBqFIY6gYi0FJH/s5/2vxGRCqqxIhIpItu8nrj729+niMjn9rHrRKSyG/g24Eb72Gl2XvtE5N/t71qIyP/YcxvsE5GR9vdbRKSniMwFmtt2/NXe95P9+Yb3E75dk7lXRBqLyDwRyRBrnoAnXLjlc2xBNxHpLdZcJHvsz1/bbynPAUbatoy0bV9pl7PHlx8NhgrUtn66WcziawEuYYm4ZQIbsFQErrL3tcV6s9RTI/7J/vwP4Dl7vTHQyk67DWhhfz8D+IOP8lZhz10B3Ad8gSWo9w3QAkua+lsgEbgXWO51bGv7cwvW07tjk1caj413A6vt9aZYSp7NgfHA7+3vmwG7gE4+7PzJ6/zWAXfY21cBTez1wcBb9vpYYLHX8X8C/s1eb4Ol+9Sitn9vs4T2EpISHgYDcE5Vu3k2RCQM+JOIDMCSo2gPtANOeR2TAay0025U1UwRGQh0AXbY8iZNsZ7EfTFPRH4PnMZS4b0N2KCWqB4i8jbQH/gbMF9EXsRqrtp+Gef1AbBIRJoBdwDbVPWc3dwVL6Uz8rUGbgKOlTu+uYhkAh2Br4BNXulXi8hNWGqgYX7KTwFSReRpezsciKZ+akAZqgkTKAx1hQexZibroaoXReR7rJucg6puswPJUOAvIjIP+AewSVVHuyjjGVVd79kQkcG+EqnqQRHpgaWZkyYiH6nqHDcnoaq/iMgWLNnrkcBaT3HAZFX9sJIszqlqNxFpDbwPTAIWYWkZfaKqd9sd/1v8HC/Avar6nRt7DQYwfRSGukNrINcOEoOA68snEJHr7TTLgf/GmhJyJ/AvIuLpc/iViNzsssxtwHD7mBZYzUbbReQ6oEhV1wDz7XLKc9Gu2fgiHUuMrT+WkB3250TPMSJys12mT1T1LDAFeNo+pjVwwt491itpIVYTnIcPgcliV69EJNFfGQaDBxMoDHWFvwI9RWQXVu3igI80yUCmiOzB6kf4s6qexrpxrhWRr7ECR4ybAlV1N1bfxZdYfRYrVHUP0BX40m4Ceg74Tx+HLwO+9nRml+MjrLmNN6s1dSdYc4lkAbtFZB/wKpXU+G1b9mLJar+EVbvZgdV/4eEToIunMxur5hFm27bP3jYYAmKGxxoMBoMhIKZGYTAYDIaAmEBhMBgMhoCYQGEwGAyGgJhAYTAYDIaAmEBhMBgMhoCYQGEwGAyGgJhAYTAYDIaA/D9Qfv4N03rSQQAAAABJRU5ErkJggg==\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
@@ -613,7 +478,9 @@
{
"cell_type": "code",
"execution_count": 9,
- "metadata": {},
+ "metadata": {
+ "collapsed": false
+ },
"outputs": [],
"source": [
"bag_clf = BaggingClassifier(\n",
@@ -624,19 +491,10 @@
{
"cell_type": "code",
"execution_count": 10,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "0.9790209790209791"
- ]
- },
- "execution_count": 10,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"bag_clf.fit(X_train, y_train)\n",
"y_pred = bag_clf.predict(X_test)\n",
@@ -1177,45 +1035,10 @@
{
"cell_type": "code",
"execution_count": 11,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"from sklearn.ensemble import AdaBoostClassifier\n",
"\n",
@@ -1244,29 +1067,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "## Additive boosting for Regression\n",
- "\n",
- "Here we present [Drucker's AdaBoost](https://pdfs.semanticscholar.org/8d49/e2dedb817f2c3330e74b63c5fc86d2399ce3.pdf) tailored for regression.\n",
- "\n",
- "In bagging, each training example is equally likely to be\n",
- "picked. In boosting, the probability of a particular\n",
- "example being in the training set of a particular machine\n",
- "depends on the performance of the prior machines on\n",
- "that example. The following is a modification of\n",
- "Adaboost by Drucker.\n",
- "\n",
- "Start by selecting a set of training data $n$ and assign to each entry a weight $w_i=1$ for $i=1,2,\\dots,n$. As we have done earlier, we could pick say $80\\%$ of the data set for training. The algorithm runs as follows:\n",
- "1. We define the probability that the training sample $i$ is in the set by $p_i = w_i/\\sum_iw_i$. We pick $n$ samples (with replacement) to form our training set. We pick a number uniformly in the range $[0,\\sum_iw_i]$.\n",
- "\n",
- "2. We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.\n",
- "\n",
- "3. Using every member of the training set with the chosen regression machine we obtain then a prediction $\\tilde{y}_i$.\n",
- "\n",
- "4. We calculate then the loss function $L_i$ for each training sample. We can use various types of loss function as long as we have a value\n",
- "\n",
- "$L_i\\in [0,1]$. \n",
- "\n",
- "## Gradient boosting: Basics with Steepest Descent\n",
+ "## Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent\n",
"\n",
"Gradient boosting is again a similar technique to Adaptive boosting,\n",
"it combines so-called weak classifiers or regressors into a strong\n",
@@ -1413,10 +1214,10 @@
"\n",
"b. fit the so-called base-learner to the negative gradient $h_m(u_m,x)$;\n",
"\n",
- "c. update the estimate $f_m(x) = f_{m-1}(x)+\\nu h_m(u_m,x)$;\n",
+ "c. update the estimate $f_m(x) = f_{m-1}(x)+h_m(u_m,x)$;\n",
"\n",
"\n",
- "4. The final estimate is then $f_M(x) = \\sum_{m=1}^M\\nu h_m(u_m,x)$.\n",
+ "4. The final estimate is then $f_M(x) = \\sum_{m=1}^M h_m(u_m,x)$.\n",
"\n",
"## Gradient Boosting, Examples of Regression"
]
@@ -1424,68 +1225,10 @@
{
"cell_type": "code",
"execution_count": 12,
- "metadata": {},
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
- " return f(**kwargs)\n",
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
- " return f(**kwargs)\n",
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
- " return f(**kwargs)\n",
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
- " return f(**kwargs)\n",
- "/Users/MortenImac/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
- " return f(**kwargs)\n"
- ]
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Max depth: 1\n",
- "Error: 0.5692982587408999\n",
- "Bias^2: 0.2744795520281314\n",
- "Var: 0.2948187067127685\n",
- "0.5692982587408999 >= 0.2744795520281314 + 0.2948187067127685 = 0.5692982587408999\n",
- "Max depth: 2\n",
- "Error: 0.5623134325774894\n",
- "Bias^2: 0.2741269616463988\n",
- "Var: 0.2881864709310905\n",
- "0.5623134325774894 >= 0.2741269616463988 + 0.2881864709310905 = 0.5623134325774893\n",
- "Max depth: 3\n",
- "Error: 0.5623132958108783\n",
- "Bias^2: 0.27412708964346316\n",
- "Var: 0.2881862061674152\n",
- "0.5623132958108783 >= 0.27412708964346316 + 0.2881862061674152 = 0.5623132958108783\n",
- "Max depth: 4\n",
- "Error: 0.5623132958108783\n",
- "Bias^2: 0.27412708964346316\n",
- "Var: 0.2881862061674152\n",
- "0.5623132958108783 >= 0.27412708964346316 + 0.2881862061674152 = 0.5623132958108783\n",
- "Max depth: 5\n",
- "Error: 0.5623132958108783\n",
- "Bias^2: 0.27412708964346316\n",
- "Var: 0.2881862061674152\n",
- "0.5623132958108783 >= 0.27412708964346316 + 0.2881862061674152 = 0.5623132958108783\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
@@ -1545,56 +1288,10 @@
{
"cell_type": "code",
"execution_count": 13,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(426, 30)\n",
- "(143, 30)\n",
- "[1. 0.86666667 0.93333333 1. 1. 0.92857143\n",
- " 1. 0.92857143 0.92857143 0.92857143]\n",
- "Test set accuracy with Random Forests and scaled data: 0.97\n"
- ]
- },
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAATgAAAEWCAYAAADy2YssAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3deZwU5Z3H8c93QBSRUxA5BE9EMR4RvKJGY7IBj2CixnjgnajxiDnX3USNxmxMolmzWaMSY0wk8Uq8FTExIYqrQURRUUFUUC4BDwREkeG3f1QN9owz010wPT1d8337qpddVU8/z1PdzW+ep6qepxQRmJnlUU2lK2BmVi4OcGaWWw5wZpZbDnBmllsOcGaWWw5wZpZbDnBtgKSJkk5LXx8n6cEWzn9LSSGpY0vmW2LZnSXdI2mppNvWI58W/1wqQdJ4SSdWuh7tRbsIcJJmS3pDUpeCbadJmljBajUqIv4YEf/W2uVKOlbSFEnLJS1I/yHu2wJZHwn0BTaNiKPWNZNyfS6SDkiD/+0Ntu+Sbp9YYj4/lDSuWLqIGBURv1/H6lpG7SLApToC31jfTJTI1ecm6VvAlcB/kQSjQcCvgdEtkP1gYGZErG6BvMplMbCPpE0Ltp0IzGypAvL4u6kKEZH7BZgNnA+8BfRIt50GTCxIsw/wBLA0/f8+BfsmAj8GHgVWAtsCAXwdeAlYBvwI2AZ4DHgXuBXolL6/J3AvyT+kt9PXAxvkf1r6+iRgUvr6e8DyguVD4IZ0X3fgt8ACYB5wKdAh3dcBuBxYArwCnJXWt2Mjn033NO+jmvn8NiQJgPPT5Upgw3TfAcBc4NvAorQ+J6f7LgZWpfVeDpwK/BAYV5D3loV1S4//lfQzfRU4ruHnUuL39aP0+1oGPAj0buLY6up/DXBWwec3F7iQ+r+RXwKvp9/vk8B+6faRDY5zWjO/m8Lv+mrgzwX5/xR4CFCl/83kZal4BVrlIJMA91ngduDSdNvaAAf0Igk8Y0haesek65um+ycCrwHD0v0bpP8o7wa6pds/SH+cW6dB43ngxPT9mwJHABsDXYHbgDsL6lf4o6/3D7kgzRYkweXgdP1O4FqgC7AZMBk4Pd13BvBi+p5ewD9oOsCNBFY3tq8gzSXA42k5fYD/A36U7jsgff8l6edyMPAe0DPd/0PqB7SG61vW1S09lneB7dN9/YBhDT+XEr+vl4EhQOd0/bImju0AkmC2D/CvdNvBwAQ+/kfw+PS77EgS0BcCGzV2XM38bgq/641JWoknAfuR/EEa2NT34CX70t6azBcC50jq02D7IcBLEXFjRKyOiJtIAsRhBWluiIjp6f4P020/jYh3I2I68BzwYES8EhFLgfHAbgAR8WZE/CUi3ouIZSR/1T9daqUldSYJaL+MiPsl9QVGAedFxIqIWAT8N/CV9C1fBq6MiNcj4i3gJ81kvymwJJrvQh4HXBIRiyJiMUnLbEzB/g/T/R9GxP0krZjtSz2+BtYAO0nqHBEL0s+2oVK+r99FxMyIWEnSmt61uUIj4v+AXpK2B04A/tBImnHpd7k6Iq4gadkWO87Gfjd1+b1HEjR/AYwDzomIuUXyswzaVYCLiOdIuofnN9jVH5jTYNscYEDB+uuNZPlGweuVjaxvAiBpY0nXSpoj6V3gYaCHpA4lVv23wIyI+Gm6PpikNbBA0juS3iFpzW1WcDyF9W14bIXeBHoXucLa8POZk25bm0eDAPke6bFnERErgKNJWqALJN0naWgJ9amrU+H3tXAd6nMjcDZwIHBHw52Svi3phfSK8DskLfXeRfJs7HezVkRMJumSiyQQWwtqVwEudRHwVer/Y5hPEjQKDSI5t1VnfaZd+TbJX/o9I6IbsH+6XcXeKOn89L2nFmx+naRL3DsieqRLt4gYlu5fQNI9rTOomSIeA94HDm8mTcPPZ1C6bV2sIOma1dm8cGdETIiIz5F0T18EflNCferqNK+RtFncSHJe9f60dbWWpP2AfydpHfeMiB4k5//qvsOmfh/N/m4knUXSEpxPcs7VWlC7C3ARMQu4BTi3YPP9wJD0VomOko4GdiRp7bWEriQtunck9SIJskVJGpXW8/C0q1V3DAtITpxfIambpBpJ20iq6/beCpwraaCknny8xUpBXktJuu5XSTo8bW1uIGmUpJ+lyW4CfiCpj6Teafqit0Q04Wlgf0mDJHUH/qPgePtK+kJ6O88HJF3d2kbyKMv3FRGvkpw6+H4ju7uSnGtcDHSUdCHJ+dc6bwBbZrlSKmkIycWh40m6/N+T1GxX2rJpdwEudQnJCW0gOUcGHErS0nqT5C/poRGxpIXKu5LkZPcSkpP1D5T4vqNJTuq/kN6ftlzSNem+E4BOJBcz3gb+TNLqgaTVMwGYBkwlubjSpIj4BfAt4Ack/4BfJ+mq3ZkmuRSYAjwDPJvmeWmJx9CwrL+S/IF5huRKZGFQqiH5DuaTXPH+NEmLqmEeZfu+ImJSRDTWOp1Acl51Jkl3+H3qdz/rbmJ+U9LUYuWkpwTGkZzHnRYRLwH/CdwoacP1OQb7iCI84aWZ5VN7bcGZWTvgAGdmueUAZ2a55QBnZrnV6tPnNEcbbBzaqEelq2EZ7Lxdv+KJrM14/bU5vLlkSdH7L5vTodvgiNUriycEYuXiCRExcn3KWx9tK8Bt1IMNd/tqpathGTw0/oJKV8EyOGi/Pdc7j1i9kg23/3JJad9/+qpiIz3Kqk0FODOrBoIqmfnJAc7MshFQU+ow6spygDOz7LRep/FajQOcmWXkLqqZ5ZlbcGaWS8ItODPLK7kFZ2Y55quoZpZPvshgZnkl3EU1sxxzC87M8sldVDPLKwEdfJHBzPLK5+DMLJ/cRTWzPHMLzsxyyy04M8sleaiWmeWZh2qZWT75IoOZ5Zm7qGaWS54Pzszyy11UM8szX2Qws9zyOTgzyyW5i2pmeeYWnJnllRzgzCyPkhnLHeDMLI8kVOMAZ2Y55RacmeWWA5yZ5ZYDnJnlk9KlCjjAmVkmQm7BmVl+1dR4JIOZ5VS1tOCqIwybWduhDEuxrKReku6QtELSHEnHNpFOki6VNE/SUkkTJQ0rlr8DnJllJqmkpQRXAauAvsBxwNVNBK6jgFOA/YBewGPAjcUyd4Azs0zqLjKsb4CT1AU4ArggIpZHxCTgbmBMI8m3AiZFxCsRUQuMA3YsVlcHODPLTDUqaSliCFAbETMLtk0DGmvB3QxsK2mIpA2AE4EHihXgiwxmlo0yXWToLWlKwfrYiBibvt4EWNog/VKgayP5LAAeAWYAtcDrwGeKFe4AZ2aZZQhwSyJieBP7lgPdGmzrBixrJO1FwAhgC2AhcDzwd0nDIuK9pgp3F9XMMmuhiwwzgY6StivYtgswvZG0uwC3RMTciFgdETcAPSlyHs4BzswyaamLDBGxArgduERSF0mfAkbT+NXRJ4CjJPWVVCNpDLABMKu5MtxFNbPsWu4+368D1wOLgDeBMyNiuqRBwPPAjhHxGvBTYDPgaaALSWA7IiLeaS5zBzgzy0YtN1QrIt4CDm9k+2skFyHq1t8HzkqXkjnAmVlm1TJUywHOzLKrjvjmiwzr43N7bMu0cefw3J/O5TvH7fux/T022YhbLv0Kk393Jo9c+1V23GqzevtrasRj153BXy5rdPidlcFDf53AnrsNY8TOQ/nlFT/72P6I4D++cx4jdh7K/nvuxrSnp67dt/Sddzj5uKPZa7ed2PuTn+CJfz3WmlVvU1pwqFZZlTXASRopaYakWZLOL2dZra2mRlz5zUMY/d1x7HbCVRx10CcYOrhPvTTfG7M/02YtZI+Tr+bUH9/B5eeOqrf/7CP3Ysacxa1Z7XattraWf//Wudxy+z08OuUZbr/tZma88Hy9NH978AFeeXkWk6e9wC9+dTXfPe/stfv+83vf5DOf+zcef+o5/vn4kwzZfofWPoQ2odTglusAJ6kDyUDaUST3qhwjqejYsWoxYocBvDzvLWYveJsPV9dy20PPcei+Q+ulGbplHyY++QoAM19bwuDNe7BZzy4ADOjTjZF7D+F39039WN5WHlOnTGarrbdhy622plOnTnzxyKMZf9899dKMv/duvnzM8Uhi+B57sXTpUhYuXMCyd9/lsUcncfyJpwDQqVMnuvfoUYnDaBPafYAD9gBmpYNjV5GMJRtdxvJaVf/e3Zi76KNRJvMWL2VAn/ojTJ6dtZDR+yd/5YfvMIBBfbszoE9y4/bPzxnJ969+kDVrovUq3c4tmD+f/gMHrl3vP2AAC+bPq59mwXwGFKbpn6SZPfsVNu3dm3POOJUD9xnON876GitWrGi1urc1LTQWtezKGeAGkIwXqzM33VaPpK9JmiJpSnzY5IiLNqexP07RIFZd/sdJ9Ojamcd/ewZnfmlPpr20kNW1axi19xAWvb2Cp2YuaJ3KGpCcX2uoYSujqTSrV6/mmaef4uTTTucf/zeFLht34X8aOYfXXlRLC66cV1EbO7qP/XrSgbdjAWq69q+a5sy8xe8ycLPua9cH9OnO/CX1h9Ate+8DTr/szrXrL95yHrMXvMNRB32CQz+1PSP32o4NO3WkW5cNuf4HX+KUS29vtfq3R/0HDGD+3Llr1+fPm8fm/frXT9N/APMK08xP0kii/4CB7D5iTwAOO/wIfvmLdhrgsg22r6hytuDmkgyMrTMQmF/G8lrVlBfns+3AXgzu14MNOnbgqIN24r5HX6yXpvsmG7FBxw4AnHzo7kyaNodl733AhWP/xrZH/oKhR1/JCRf/mYlTX3VwawW77T6CV16exZzZr7Jq1Sru+PMtjDz40HppRh5yGLfeNI6IYMrkx+nWrRubb96Pvn03Z8CAgbw0cwYAD0/8O9sPbacXGUh6MKUslVbOFtwTwHaStgLmAV8BcnM/RG3tGr555f3cc/kYOtTU8Pv7n+KF2Ys57QvJxAnX3T2FoYN7c933v0Rt7RpenLOYMy67q8K1bt86duzIZVf8kqMOP4Q1tbUcO+Ykhu44jN9ddy0AJ592Op/7/Cj+NmE8I3YeSufOnfmfa65b+/6fXHElZ5x6Ah+uWsXgrbbmV1df11RROdc2up+lUGPnHFosc+lg4EqgA3B9RPy4ufQ1XfvHhrt9tWz1sZY3d/wFla6CZXDQfnvy9NQn1ys6bbT5kBh84q9KSjvzZyOfbGa6pLIr60iGiLgfuL+cZZhZK2sj3c9SeKiWmWUikhvdq4EDnJll5hacmeVWtVxkcIAzs2x8Ds7M8kqoxSa8LDcHODPLzC04M8stn4Mzs3zyOTgzy6tkLGp1RDgHODPLrErimwOcmWXnkQxmlk9VNB+cA5yZZVI3H1w1cIAzs4yqZz44Bzgzy6xK4psDnJllJF9kMLOc8n1wZpZrDnBmlltVEt8c4MwsO7fgzCyfPNjezPIqmfCyOiKcA5yZZVZTJU246ph32MzaFKm0pXg+6iXpDkkrJM2RdGwzabeWdK+kZZKWSPpZsfzdgjOzTNSyg+2vAlYBfYFdgfskTYuI6fXLVCfgr2n6o4FaYEixzB3gzCyzljgFJ6kLcASwU0QsByZJuhsYA5zfIPlJwPyI+EXBtmeKldFkgJP0KyCa2h8R5xbL3MzyKcNFht6SphSsj42IsenrIUBtRMws2D8N+HQj+ewFzJY0HhgBPAecExHPNld4cy24Kc3sM7N2SiRXUku0JCKGN7FvE2Bpg21Lga6NpB0IHAh8AXgI+AZwl6ShEbGqqcKbDHAR8fvCdUldImJFU+nNrP1oobtElgPdGmzrBixrJO1KYFJEjAeQdDnwA2AHklZf4/UsVgNJe0t6HnghXd9F0q9Lqr6Z5Y+S+eBKWYqYCXSUtF3Btl2A6Y2kfYZmTpk1pZTbRK4EPg+8CRAR04D9sxZkZvnREreJpD3C24FLJHWR9ClgNHBjI8nHAXtJ+qykDsB5wBLShldTSroPLiJeb7CptpT3mVn+iORG31KWEnwd6AwsAm4CzoyI6ZIGSVouaRBARMwAjgeuAd4mCYRfaO78G5R2m8jrkvYBIr0X5VyKRE0zy7eWGqoVEW8Bhzey/TWSixCF224nafGVrJQW3BnAWcAAYB7JzXhnZSnEzPKj1O5pWxjNVbQFFxFLgONaoS5mViVyMxY1Hf91j6TFkhZJukvS1q1ROTNrm1TiUmmldFH/BNwK9AP6A7eRnAw0s3aqhW4TKbtSApwi4saIWJ0u41iH+1HMLB+Sq6ilLZXW3FjUXunLf0g6H7iZJLAdDdzXCnUzs7ZI+Zjw8kmSgFZ3JKcX7AvgR+WqlJm1bW2h+1mK5saibtWaFTGz6lDXRa0GJc0HJ2knYEdgo7ptEfGHclXKzNq2qm/B1ZF0EXAASYC7HxgFTAIc4MzaqeoIb6VdRT0SOAhYGBEnk4z237CstTKzNkuCDjUqaam0UrqoKyNijaTVkrqRDIr1jb5m7VhuuqjAFEk9gN+QXFldDkwua63MrE2rkvhW0ljUr6cvr5H0ANAtIoo+7MHM8kmUPBVSxTV3o+8nm9sXEVPLUyUza9PayEwhpWiuBXdFM/sC+EwL14XdhvTn0b9f3NLZWhn1HHF2patgGXwwo+Hcteum6s/BRcSBrVkRM6sOAjpUe4AzM2tKG7gDpCQOcGaWmQOcmeVSMh15dUS4Umb0laTjJV2Yrg+StEf5q2ZmbVW1zAdXylCtXwN7A8ek68uAq8pWIzNr83Lz0Blgz4j4pKSnACLi7fTxgWbWDgno2BaiVwlKCXAfpk+SDgBJfYA1Za2VmbVpVRLfSgpw/wPcAWwm6ccks4v8oKy1MrM2S6U/tb7iShmL+kdJT5JMmSTg8Ijwk+3N2rEqiW8lTXg5CHgPuKdwW0S8Vs6KmVnb1RaukJailC7qfXz08JmNgK2AGcCwMtbLzNooQZuYzLIUpXRRP1G4ns4ycnoTyc0s79rIPW6lyDySISKmShpRjsqYWXVQlTyVoZRzcN8qWK0BPgksLluNzKxNy9tjA7sWvF5Nck7uL+WpjplVg1wEuPQG300i4rutVB8zqwLVMti+uSnLO0bE6uamLjez9id5bGCla1Ga5qpZ9+SspyXdLWmMpC/VLa1ROTNrm2rS0QzFlmIk9ZJ0h6QVkuZIOraE9/xdUkgqeoqtlHNwvYA3SZ7BUHc/XAC3l/BeM8uZFr7IcBWwCugL7ArcJ2laRExvtGzpODLc/dFcws3SK6jP8VFgqxOlFmBm+dMSp+AkdQGOAHaKiOXAJEl3A2OA8xtJ3x24CDgBeKyUMpoLcB2ATaDRG14c4MzaLVFT+n1wvSVNKVgfGxFj09dDgNqImFmwfxrw6Sby+i/gamBhqYU3F+AWRMQlpWZkZu2DyNSCWxIRw5vYtwmwtMG2pdS/NS0pUxoOfAr4BjCw1MKbC3DVcR3YzFqXoGPLnIRbDnRrsK0byazhHxUn1ZDMLP6N9M6Okgto7irqQSXnYmbtRl0LrgWmLJ8JdJS0XcG2XYCGFxi6AcOBWyQtBJ5It8+VtF9zBTT34Oe3ilbPzNqllpjwMiJWSLoduETSaSRXUUcD+zRIuhToX7C+BcltbLtTZNioHxtoZpm14ECGrwPXA4tIbkc7MyKmp/NQPg/smM49ufbCgqSN0pdvRMTq5jJ3gDOzTERpj+MrRdpTPLyR7a+RXIRo7D2zKfEagQOcmWWjlumitgYHODPLJBnJ4ABnZjlVHeHNAc7M1kGVNOAc4MwsK1X/fHBmZo1pyauo5eYAZ2aZ+SKDmeWTcjBluZlZY9xFNbNccwvOzHKrOsKbA5yZZSSgg1twZpZXVRLfHODMLCuhKumkOsCZWWZuwZlZLiW3iVRHhHOAM7NsSnveQpvgAGdmmXmolpnlUjLhZaVrURoHODPLzFdRzSy3qqSHWjVjZtukByc8wM7DtmfY0G35+c8u+9j+iOBb553LsKHbMmK3nXlq6tS1+04/7RQG9d+M3XfdqTWr3O59bp8dmHbHBTx310V85+TPfWx/j66dueWKrzL5lv/gkRu/w47b9Fu776xjDmDKbf/Jk3/+Pmcfe0Ar1rrtUYn/VVrZApyk6yUtkvRcucqopNraWs479yzuumc8Tz3zPLfdfBMvPP98vTQTHhjPy7Ne4rkXXuJ/rx7LuWefuXbfmBNP4q57H2jtardrNTXiyvO/zOizf81uR1zKUSN3Z+jWm9dL871TP8+0GXPZ4+ifcOoFN3L5d48EYMdt+nHyl/ZhvzE/Z4+jf8Ko/Xdim0F9KnEYFVd3Dq6UpdLK2YK7ARhZxvwr6onJk9lmm23Zauut6dSpE0cd/RXuveeuemnuvfsujj3+BCSx5157sXTpOyxYsACAfffbn169elWi6u3WiJ225OXXlzB73pt8uLqW2yZM5dADdq6XZujWmzNx8gwAZs5+g8H9e7FZr64M3WpzJj87m5Xvf0ht7RoeeXIWow/cpRKHUXkSNSUulVa2ABcRDwNvlSv/Sps/fx4DB26xdn3AgIHMmzevaJr5DdJY6+m/WXfmvvH22vV5b7zNgD7d66V5duY8Rh+0KwDDhw1mUL9eDOjbg+kvz2ffT25Lr+5d6LzRBozcdxgDN+/ZqvVvS1TiUmkVv8gg6WvA1wC2GDSowrUpXUR8bFvDObJKSWOtp7FzQg2/oct/91cu/+6RPH7z+Ux/aT7TZsxlde0aZrz6Blfc8FfuvfpsVqz8gGdmzmP16trWqXgb4+eiZhARY4GxALvvPvzjEaGNGjBgIHPnvr52fd68ufTv379omn4N0ljrmbfoHQb2/ajVNaBvT+YvXlovzbIV73P6D8etXX/xvouZPe9NAH5/52P8/s7HALj47MOY98Y7rVDrtqk6wpuvoq6z4SNGMGvWS8x+9VVWrVrFbbfczCGHfqFemkMO+wJ/GvcHIoJ/Pf443bp1p1+/fk3kaOU2Zfocth3Uh8H9N2WDjh046vOf5L6Jz9RL032TzmzQsQMAJ39xHyZNncWyFe8D0KfnJgBssXlPRn9mF259YErrHkBbUiV91Iq34KpVx44d+e9f/i+HHfJ5amtrOfGkU9hx2DB+c+01AHz19DMYOepgJoy/n2FDt2Xjzhtz7XW/W/v+E44/hkf+OZElS5awzZYDueDCiznplFMrdTjtQm3tGr7501u559dn0aFG/P6ux3nhlYWcduS+AFz350kM3XpzrvvRGGpr1/DiKws54+I/rn3/TZefRq8eXfhwdS3nXXYr7yxbWalDqbhq6aKqsfNELZKxdBNwANAbeAO4KCJ+29x7dt99eDz6r3b8V7EK9RxxdqWrYBl8MONW1ry3aL2i0w6f2C3+cNfEktLusU2PJyNi+PqUtz7K1oKLiGPKlbeZVVh1NODcRTWzbJLTa9UR4RzgzCwbzwdnZnlWJfHNt4mYWVZCKm0pmpPUS9IdklZImiPp2CbSnSjpSUnvSpor6WeSijbQHODMLDOptKUEVwGrgL7AccDVkoY1km5j4DySuzL2BA4CvlMsc3dRzSyTlrqHV1IX4Ahgp4hYDkySdDcwBji/MG1EXF2wOk/SH4EDi5XhFpyZZdcyIxmGALURMbNg2zSgsRZcQ/sD04slcgvOzDLLcJtIb0mFd++PTcefA2wCLG2QfinQtdmypZOB4cBpxQp3gDOzzDLcJrKkmZEMy4FuDbZ1A5Y1Xa4OBy4DPhsRS4oV7i6qmWVT4gWGEoLgTKCjpO0Ktu1CE11PSSOB3wCHRcSzpVTVAc7MMmuJZzJExArgduASSV0kfQoYDdz4sfKkzwB/BI6IiMml1tMBzswyES16m8jXgc7AIuAm4MyImC5pkKTlkupmwb0A6A7cn25fLml8scx9Ds7MMmupkQwR8RZweCPbXyO5CFG3XvSWkMY4wJlZdlUyVssBzswyq5YJLx3gzCyz6ghvDnBmti6qJMI5wJlZJp7w0szyyxNemlmeVUl8c4Azs6xKm8yyLXCAM7PMqiS+OcCZWTZt5KH1JXGAM7PsqiTCOcCZWWa+TcTMcsvn4MwsnwQ1DnBmll/VEeEc4Mwsk7oJL6uBA5yZZVYl8c0BzsyycwvOzHLLQ7XMLLeqI7w5wJlZRhmemFVxDnBmlplHMphZflVHfHOAM7PsqiS+OcCZWVbyYwPNLJ+qaSRDTaUrYGZWLm7BmVlm1dKCc4Azs8x8m4iZ5ZNv9DWzvKqmiwwOcGaWmbuoZpZbbsGZWW5VSXxzgDOzdVAlEc4BzswyEVTNUC1FRKXrsJakxcCcStejDHoDSypdCcskr9/Z4Ijosz4ZSHqA5PMpxZKIGLk+5a2PNhXg8krSlIgYXul6WOn8neWDx6KaWW45wJlZbjnAtY6xla6AZebvLAd8Ds7McsstODPLLQc4M8stB7gykjRS0gxJsySdX+n6WHGSrpe0SNJzla6LrT8HuDKR1AG4ChgF7AgcI2nHytbKSnADULEbU61lOcCVzx7ArIh4JSJWATcDoytcJysiIh4G3qp0PaxlOMCVzwDg9YL1uek2M2slDnDl09hoZN+TY9aKHODKZy6wRcH6QGB+hepi1i45wJXPE8B2kraS1An4CnB3hetk1q44wJVJRKwGzgYmAC8At0bE9MrWyoqRdBPwGLC9pLmSTq10nWzdeaiWmeWWW3BmllsOcGaWWw5wZpZbDnBmllsOcGaWWw5wVURSraSnJT0n6TZJG69HXjdIOjJ9fV1zEwFIOkDSPutQxmxJH3v6UlPbG6RZnrGsH0r6TtY6Wr45wFWXlRGxa0TsBKwCzijcmc5gkllEnBYRzzeT5AAgc4AzqzQHuOr1CLBt2rr6h6Q/Ac9K6iDp55KekPSMpNMBlPhfSc9Lug/YrC4jSRMlDU9fj5Q0VdI0SQ9J2pIkkH4zbT3uJ6mPpL+kZTwh6VPpezeV9KCkpyRdSwnPP5d0p6QnJU2X9LUG+65I6/KQpD7ptm0kPZC+5xFJQ1viw7R88pPtq5CkjiTzzD2QbtoD2CkiXk2DxNKIGCFpQ+BRSQ8CuwHbA58A+gLPA9c3yLcP8Btg/zSvXhHxlqRrgOURcXma7k/Af0fEJEmDSEZr7ABcBEyKiEskHQLUC1hNOCUtozPwhKS/RMSbQBdgakR8W2So3agAAAHGSURBVNKFad5nkzwM5oyIeEnSnsCvgc+sw8do7YADXHXpLOnp9PUjwG9Juo6TI+LVdPu/ATvXnV8DugPbAfsDN0VELTBf0t8byX8v4OG6vCKiqXnRPgvsKK1toHWT1DUt40vpe++T9HYJx3SupC+mr7dI6/omsAa4Jd0+Drhd0ibp8d5WUPaGJZRh7ZQDXHVZGRG7Fm5I/6GvKNwEnBMRExqkO5ji0zWphDSQnNrYOyJWNlKXksf+STqAJFjuHRHvSZoIbNRE8kjLfafhZ2DWFJ+Dy58JwJmSNgCQNERSF+Bh4CvpObp+wIGNvPcx4NOStkrf2yvdvgzoWpDuQZLuImm6uoDzMHBcum0U0LNIXbsDb6fBbShJC7JODVDXCj2WpOv7LvCqpKPSMiRplyJlWDvmAJc/15GcX5uaPjjlWpKW+h3AS8CzwNXAPxu+MSIWk5w3u13SND7qIt4DfLHuIgNwLjA8vYjxPB9dzb0Y2F/SVJKu8mtF6voA0FHSM8CPgMcL9q0Ahkl6kuQc2yXp9uOAU9P6TcfTwFszPJuImeWWW3BmllsOcGaWWw5wZpZbDnBmllsOcGaWWw5wZpZbDnBmllv/D5RuXA3+U8upAAAAAElFTkSuQmCC\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
@@ -1664,105 +1361,10 @@
{
"cell_type": "code",
"execution_count": 14,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 0\n",
- "Error: 0.3122934461614368\n",
- "Bias^2: 0.31229345725057706\n",
- "Var: 3.552713678800501e-15\n",
- "0.3122934461614368 >= 0.31229345725057706 + 3.552713678800501e-15 = 0.3122934572505806\n",
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 1\n",
- "Error: 0.3630793080432832\n",
- "Bias^2: 0.3249573668730529\n",
- "Var: 0.038121938705444336\n",
- "0.3630793080432832 >= 0.3249573668730529 + 0.038121938705444336 = 0.3630793055784972\n",
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 2\n",
- "Error: 0.3604382790875038\n",
- "Bias^2: 0.32184437765562457\n",
- "Var: 0.03859391063451767\n",
- "0.3604382790875038 >= 0.32184437765562457 + 0.03859391063451767 = 0.36043828829014224\n",
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 3\n",
- "Error: 0.360435136192909\n",
- "Bias^2: 0.32182779880247675\n",
- "Var: 0.03860734403133392\n",
- "0.360435136192909 >= 0.32182779880247675 + 0.03860734403133392 = 0.3604351428338107\n",
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 4\n",
- "Error: 0.360435136192909\n",
- "Bias^2: 0.32182779880247675\n",
- "Var: 0.03860734403133392\n",
- "0.360435136192909 >= 0.32182779880247675 + 0.03860734403133392 = 0.3604351428338107\n",
- "[11:32:43] WARNING: /Users/travis/build/dmlc/xgboost/src/learner.cc:480: \n",
- "Parameters: { colsaobjective } might not be used.\n",
- "\n",
- " This may not be accurate due to some parameters are only used in language bindings but\n",
- " passed down to XGBoost core. Or some parameters are not used but slip through this\n",
- " verification. Please open an issue if you find above cases.\n",
- "\n",
- "\n",
- "Max depth: 5\n",
- "Error: 0.360435136192909\n",
- "Bias^2: 0.32182779880247675\n",
- "Var: 0.03860734403133392\n",
- "0.360435136192909 >= 0.32182779880247675 + 0.03860734403133392 = 0.3604351428338107\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
@@ -1823,88 +1425,11 @@
},
{
"cell_type": "code",
- "execution_count": 16,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(426, 30)\n",
- "(143, 30)\n",
- "Test set accuracy with xgboost and scaled data: 1.00\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeXhN5/bA8e8riFJjhBpiDolMhMTQgRoqLaVKDb8W1dZY3GqVXr1UdVBDq2qoqa0qLaWX0upEqdYsrphFaohZTJGIRIb1++PEaWYROTk5sT7Pcx7Z+7xn77UPsva7373Xa0QEpZRSKjOF7B2AUkqp/E0ThVJKqSxpolBKKZUlTRRKKaWypIlCKaVUljRRKKWUypImCqWUUlnSRKEcnjHmuDHmhjEm2hhzzhizwBhzf5o2zY0xvxtjoowxkcaY1caY+mnalDLGfGyMCU/eVljycvlM9muMMcOMMfuMMdeNMaeMMcuMMT62PF6l8pomClVQPCki9wMNgIbAv2+9YYxpBvwKfA9UBmoCIcAmY0yt5DZFgXWAFxAElAKaA5eAwEz2OQ34FzAMKAfUBVYC7e80eGNM4Tv9jFJ5RROFKlBE5BzwC5aEccskYKGITBORKBG5LCL/AbYC45Lb9AaqAZ1F5ICIJInIBRF5R0TWpN2PMcYdeBnoKSK/i0iciMSIyGIR+SC5zQZjzEspPvO8MeavFMtijHnZGHMEOGKMmW2MmZJmP98bY15N/rmyMeY7Y0yEMeaYMWZYinaBxpidxphrxpjzxpiP7uJrVCoVTRSqQDHGVAUeB8KSl4tj6Rksy6D5t0Db5J/bAD+LSHQ2d9UaOCUi2+8uYp4CmgD1ga+B7sYYA2CMKQs8BiwxxhQCVmPpCVVJ3v8rxph2yduZBkwTkVJA7eRjUypXaKJQBcVKY0wUcBK4ALyVvL4cln/nZzP4zFng1viDSyZtMnOn7TMzIbmHcwP4ExDg4eT3ugJbROQMEAC4ish4EbkpIkeBeUCP5LbxQB1jTHkRiRaRrbkQm1KAJgpVcDwlIiWBloAH/ySAK0ASUCmDz1QCLib/fCmTNpm50/aZOXnrB7FU6FwC9Exe9X/A4uSfqwOVjTFXb72A0UDF5PdfxDJGcsgYs8MY0yEXYlMK0EShChgR+QNYAExJXr4ObAGeyaB5NywD2ABrgXbGmBLZ3NU6oKoxpnEWba4DxVMsP5BRyGmWvwG6GmOqY7kk9V3y+pPAMREpk+JVUkSeABCRIyLSE6gATASW38GxKJUlTRSqIPoYaGuMuTWg/QbQJ/lW1pLGmLLGmHeBZsDbyW2+wvLL+DtjjIcxppAxxsUYM9oY80TaHYjIEWAW8I0xpqUxpqgxppgxpocx5o3kZruBp40xxY0xdbCc9WdJRP4HRADzgV9E5GryW9uBa8aYUcaY+4wxTsYYb2NMAIAx5jljjKuIJAG3PpN4J1+aUpnRRKEKHBGJABYCY5KX/wLaAU9jGVc4geUW2oeSf+EjInFYBrQPAb8B17D8ci4PbMtkV8OAGcBMLL+c/wY6Yxl0BpgK3ATOA1/yz2Wk2/kmOZavUxxTIvAklru5jmG5ZDYfKJ3cJAjYb4yJxjKw3UNEYrO5P6WyZHTiIqWUUlnRHoVSSqksaaJQSimVJU0USimlsqSJQimlVJYcrhBZ+fLlpUaNGvYOQymlHEpwcPBFEXHNyWcdLlHUqFGDnTt32jsMpZRyKMaYEzn9rF56UkoplSVNFEoppbKkiUIppVSWNFEopZTKkiYKpZRSWdJEoZRSKks2SxTGmM+NMReMMfsyed8YYz4xxoQZY/YYY/xtFYtSSqmcs+VzFAuwlGBemMn7jwPuya8mwKfJf2YpOPgMxrxtXRZ5C1xnpG4UMYS5c4MZMOAH66p+/fyZ+1BNeG3DP+161YePWtGo0Vx27fpnVsudO/vRaOSfsCfin7ZruxGckEDjxvOsq/z9KxH8+VPQJsX0xL6usK47/fuvZt68XdbVc+Z0oP+hSPjqwD9tP2wJvb1THY8ekx6THpMeU24f081X13I3bNajEJGNwOUsmnQCForFVqCMMeaOp5Y060yG6wYcHJBq3bzT8+ifZt3c0/Mw6wy7ooJTrW+8vTHBadY12t6YxttTT2a2KyqYRmnWBUcFY9YZ5p2el2r9gIMDmJtmXf+DAzKNX49Jj0mPSY8pN46p+cInaXR1Urr2d8KeYxRVSDFfMHAqeV06xpj+xpidxhh9JFsppe5AuQolOJB47q62Yc9EkT4lpp8/2LJSZK6INBaRdPMTS+v0H5HWwhzPOanW9avSj7lp1vWv0g9pLfiXbJRq/c7AnTRKsy44cCc7A1PnKf+SjQhOs65RyUZIa6FflX6p1s/xnEP/NOvmes7JNH49Jj0mPSY9ppwdUySww3pMPwQtIazs2HTt74RNZ7gzxtQAfhAR7wzemwNsEJFvkpcPAy1F5Gzatik1LlJNdpYZCRFDbBCxUko5poSEBD755BPGjh3L9evX2bhxIw8//LD1fWNMcEYn29lhzx7FKqB38t1PTYHI2yUJpZRS6W3bto3GjRvz2muvcf36dbp06UKtWrVybfs2u+vJGPMN0BIob4w5BbwFFAEQkdnAGuAJIAyIAfraKhallCqIrly5wujRo5kzZw4iQo0aNZgxYwbt27fP1f3YLFGISM/bvC/Ay3e63YiS0ZZb1pRS6h739ttvM3v2bAoXLsyIESMYM2YMxYsXz/X92HSMwhZMPSNy2LFiVkqp3JKQkEDhwpZz/IsXL/Liiy/y3nvv4e2dbig4lbsZo9BEoZRSDiA2NpaJEyeycuVKtm3bRtGiRe/o83eTKBxuhjullLrXrFu3jkGDBnHkyBEAfvnlF5588sk8278WBVRKqXzq/PnzPPfcc7Rp04YjR47g6enJhg0b8jRJgCYKpZTKlxYtWoSHhweLFy+mWLFivP/+++zevZsWLVrkeSx66UkppfKhpKQkrl69SlBQEDNnzszV5yLulMP1KDxPVYTWS+0dhlJK5aro6Gh+++0363KvXr347bffWLNmjV2TBDhgoih+s2jq0r5KKeXgVq5ciaenJ08++SRhYWEAGGNo06YNxmRUFi9vOVyiUEqpguLEiRN06tSJzp07c+rUKXx8fIiLi7N3WOloolBKqTwWHx/P5MmTqV+/PqtWraJkyZJMnz6drVu34uXlZe/w0nG4weyDVc7B993sHYZSSuXYsGHDmD17NgDdunVj6tSpVK5c2c5RZc7hehQxzvHgV8HeYSilVI698soreHp68tNPP7F06dJ8nSRAS3gopZRNiQiLFi1izZo1fP3119bB6aSkJAoVyrtzdUedj0IppQq0w4cP07p1a3r37s2SJUv46aefrO/lZZK4W44TqVJKOYgbN24wduxYfH19Wb9+PS4uLixYsIDHH3/c3qHliMMNZiulVH62du1aBg4cyN9//w3Aiy++yMSJE3FxcbFzZDnncD2K4nFFIOSCvcNQSqkMbd68mb///hsvLy/+/PNP5s+f79BJAhywR+F5+gFo8y1EDLF3KEopRWJiImFhYdSrVw+AUaNGUb58eV566aU7njMiv3K4HoVSSuUX//vf/2jevDkPPfQQly9fBsDZ2ZnBgwcXmCQBmiiUUuqORUVFMXz4cBo3bsz27dtxdna2jkkURA6XKGKK3gRfV3uHoZS6B4kI3333HZ6ennz88ccADB8+nIMHDxIQEGDn6GxHH7hTSqls+te//sUnn3wCQEBAAHPmzKFhw4Z2jip79IE7pZTKA507d6Z06dLMnDmTLVu2OEySuFvao1BKqUz89ddfrF+/njFjxljXXbt2jVKlStkxqpy5mx6Fw90eq5RStnbp0iVGjRrFZ599BkDr1q1p3rw5gEMmibuliUIppZKJCAsXLmTEiBFcvHiRIkWK8MYbb9wzl5gyo4lCKaWAgwcPMmjQIP744w8AHn30UWbNmoWHh4edI7M/hxvMrhZRFl793d5hKKUKmI8++og//vgDV1dXvvrqK9atW6dJIpnD9Shco+6Hrw7AR63sHYpSysFFRkZSunRpACZMmECJEiUYO3Ys5cqVs3Nk+YvD9SiUUupunTlzhu7du9O0aVNu3rwJQPny5fn44481SWRAE4VS6p6RmJjI9OnT8fDw4NtvvyU8PJxdu3bZO6x8z+ESxYnyl+HDlvYOQynlYIKDg2nSpAnDhg0jKiqKjh07cvDgQZo2bWrv0PI9myYKY0yQMeawMSbMGPNGBu+XNsasNsaEGGP2G2P63m6bF0tdh97etglYKVUgjRs3jsDAQIKDg3Fzc2PlypV8//33VKtWzd6hOQSbJQpjjBMwE3gcqA/0NMbUT9PsZeCAiPgBLYEPjTEFpzavUipfqFWrFsYYXnvtNQ4cOECnTp3sHZJDseVdT4FAmIgcBTDGLAE6AQdStBGgpDHGAPcDl4EEG8aklLoHHD16lB07dtC9e3cAevXqRZMmTayTC6k7Y8tLT1WAkymWTyWvS2kG4AmcAfYC/xKRpLQbMsb0N8bsNMbstFWwSinHd/PmTd5//328vLzo06cPYWFhABhjNEncBVsmCpPBurTV/NoBu4HKQANghjEmXSEVEZkrIo1zWtBKKVXwbdy4kQYNGvDmm28SGxtL165d78m6TLZgy0RxCnBLsVwVS88hpb7Af8UiDDgG6KOQSqlsu3jxIn379qVFixYcPHgQd3d31q5dy6JFi6hQoYK9wysQbJkodgDuxpiayQPUPYBVadqEA60BjDEVgXrA0aw22uioG7jOsEG4SilHNHDgQBYsWICzszNvv/02e/bsoXXr1vYOq0Cx2WC2iCQYY4YAvwBOwOcist8YMzD5/dnAO8ACY8xeLJeqRonIRVvFpJQqGJKSkihUyHKe+95773Hjxg0+/vhj3N3d7RxZweRwExc1LlJNdpYZCRFD7B2KUiqPxcTE8M4777B7927WrFmD5YZJlR06cZFSqsD78ccfGTJkCMePH8cYw/bt22nSpIm9w7onOFwJj+BaJ7U3odQ95NSpU3Tp0oUOHTpw/Phx/Pz82Lx5syaJPORwiUIpde+YNWsWnp6e/Pe//6VEiRJ89NFH7Ny5U+sz5TG99KSUyrcuXrxIdHQ0nTt3Ztq0abi5ud3+QyrXOdxgtqlnRA47VsxKqey5evUqhw4dsvYY4uLiWL9+PUFBQXaOzPHdzWC2XnpSStmdiLBkyRI8PT3p2LEjly9fBsDZ2VmTRD6giUIpZVdhYWEEBQXRs2dPzp07h7u7O5GRkfYOS6XgcImi/LUSsHCfvcNQSt2luLg43nnnHby9vfn1118pW7Ys8+bN488//6RmzZr2Dk+lkO3BbGNMCRG5bstgsqP6xXLw2gadvEgpB9e9e3e+//57AHr37s3kyZO1NlM+ddsehTGmuTHmAHAwednPGDPL5pEppQq0V155BQ8PD37//Xe+/PJLTRL5WHZ6FFOxlANfBSAiIcaYR2walVKqQElKSuLzzz/n4MGDfPjhhwC0bNmSffv24eTkZOfo1O1k69KTiJxMU1Ml0Tbh3F5EyWjolXZGVaVUfrV3714GDhzI5s2bActlJj8/PwBNEg4iO4PZJ40xzQExxhQ1xowg+TKUPYS7XoGPWtlr90qpbLp+/TojR46kYcOGbN68mQceeIAlS5bg6+tr79DUHcpOj2IgMA3LNKangF+BwbYMSinl2FavXs2QIUMIDw/HGMPLL7/Me++9R+nSpe0dmsqB7CSKeiLybMoVxpgHgU22CUkp5ehWrlxJeHg4DRs2ZM6cOQQEBNg7JHUXblvCwxizS0T8b7cur2gJD6Xyn4SEBE6fPk316tUBS42mJUuWMHDgQAoX1pJy+YFN5qMwxjQDmgOuxphXU7xVCsuMdUopxdatWxk4cCBxcXGEhIRQtGhRypcvz5AhOh1AQZHVYHZR4H4syaRkitc1oKvtQ1NK5WdXrlxh0KBBNG/enJCQEGJjYzl+/Li9w1I2kJ1LT9VF5EQexXNb9Us8IAeaToN13e0dilL3JBHhm2++Yfjw4Vy4cIHChQvz+uuv85///IfixYvbOzyVCVtPhRpjjJkMeAHFbq0UEbvco1r8ZlHYE2GPXSulgGeffZZvvvkGgIcffphPP/0ULy8vO0elbCk7z1EsBg4BNYG3gePADhvGpJTKx4KCgnBxceHzzz9nw4YNmiTuAdlJFC4i8hkQLyJ/iMgLgM5DqNQ9Yu3atcyZM8e63KtXL0JDQ+nbty+FCjlcAWqVA9n5W45P/vOsMaa9MaYhUNWGMWXpYJVzsLabvXav1D3j/PnzPPvss7Rt25Z//etf/P333wAYYyhXrpydo1N5KTtjFO8aY0oDrwHTsdwe+4pNo8pCjHM8+GmVSaVsJSkpiblz5/LGG28QGRlJsWLFGDt2rM5XfQ+7baIQkR+Sf4wEHgXrk9lKqQImJCSEAQMGsG3bNgAef/xxZsyYQa1atewcmbKnrB64cwK6Yanx9LOI7DPGdABGA/cBDfMmRKVUXhk5ciTbtm2jcuXKTJs2jS5dupCmcrS6B2X6HIUxZgHgBmwHmgAngGbAGyKyMq8CTBeXlvBQKteICDExMZQoUQKAw4cPM3v2bN5++21KlSpl5+hUbrqb5yiyShT7AF8RSTLGFAMuAnVE5FzOQ717miiUyh0nTpxg6NChXL9+nbVr12rPoYC7m0SR1V1PN0UkCUBEYoFQeycJgOJxRSDkgr3DUMphxcfHM2nSJOrXr8/q1avZsWMHR44csXdYKh/LKlF4GGP2JL/2pljea4zZk1cBpuV5+gFo8629dq+UQ9u0aRP+/v6MGjWKmJgYunfvzqFDh6hbt669Q1P5WFZ3PXnmWRRKKZsbOnQoM2bMAKBWrVrMnDmToKAgO0elHEGmiSI/FQJUSt09V1dXihQpwqhRoxg9ejT33XefvUNSDsKmz98bY4KMMYeNMWHGmDcyadPSGLPbGLPfGPPH7bYZU/Qm+LrmfrBKFTCHDh3i119/tS6PGjWKPXv28M4772iSUHfktmXGc7xhy3MYoUBbLHNt7wB6isiBFG3KAJuBIBEJN8ZUEJEsR6r1rielsnbjxg3ef/99Jk6cSJkyZTh06JCW3FA2LzOOMeY+oJqIHL6DbQcCYSJyNHkbS4BOwIEUbf4P+K+IhAPcLkkopbL266+/MnjwYGtdpo4dO+ptr+qu3fbSkzHmSWA38HPycgNjzKpsbLsKcDLF8qnkdSnVBcoaYzYYY4KNMb2zF7ZSKqWzZ8/So0cP2rVrx99//42Xlxd//vkn8+fPp2zZsvYOTzm47PQoxmHpHWwAEJHdxpga2fhcRqcxaa8ZFQYaAa2xlAXZYozZKiKhqTZkTH+gP2BJLUqpVJ5++mm2bt3Kfffdx7hx4xg+fDhFihSxd1iqgMjOYHaCiETmYNunsJQAuaUqcCaDNj+LyHURuQhsBPzSbkhE5opI45xeX1OqIEo5vvjBBx/QoUMHDhw4wMiRIzVJqFyVnUSxzxjzf4CTMcbdGDMdywD07ewA3I0xNY0xRYEeQNpLVt8DDxtjChtjimOpKXXwDuJX6p4TFRXF8OHDGTBggHVdixYtWL16NTVq1LBfYKrAyk6iGIplvuw44Gss5cZvOx+FiCQAQ4BfsPzy/1ZE9htjBhpjBia3OYhl7GMPluKD80VkX1bbrRZRFl79PRthK1WwiAjfffcdnp6efPzxx3zxxRccP37c3mGpe8Btb481xjQUkf/lUTy31bhINdlZZiREDLF3KErlmWPHjjFkyBDWrFkDQGBgILNnz6ZhQ632r7LHVkUBb/nIGHPIGPOOMUZnUVcqD4kIEydOxMvLizVr1lC6dGlmzZrF5s2bNUmoPHPbRCEijwItgQhgbnJRwP/YOjCllGV+6tDQUG7cuEHPnj05dOgQgwYNwsnJyd6hqXvIHT2ZbYzxAUYC3UWkqM2iyoJrpfslYuJW6O1tj90rZXMXL17k3LlzeHt7W5f/97//0bZtWztHphyZTSYuSrFxT6A70BW4BCwBvrPXU9RawkMVVCLCl19+yYgRI3B1dSUkJISiRe1yPqYKIFuX8PgC+AZ4TETSPgehlMoFBw8eZODAgWzcuBEAPz8/rly5QsWKFe0cmVLZSBQi0jQvAlHqXhQTE8N7773H5MmTiY+Px9XVlY8++ohnn31WazSpfCPTRGGM+VZEuiXPbpfyWo8BRER8bR6dUgWYiNCqVSu2bdsGwIABA5gwYYLWZlL5TlY9in8l/9khLwJR6l5jjGHw4MHExMQwZ84cmjVrZu+QlMpQdgazJ4rIqNutyys6mK0cVWJiIrNmzSI+Pp5XX30VsPQqEhIStDaTsjlbP3CX0T15j+dkZ7mh0VE3cJ1hr90rlSM7d+6kSZMmDBs2jNGjR3PmjOW+EGOMJgmV72WaKIwxg5LHJ+oZY/akeB3DUptJKXUbkZGRDB06lMDAQIKDg3Fzc2Pp0qVUrlzZ3qEplW1ZjVF8DfwETABSzncdJSKXbRqVUg5ORFi2bBmvvPIKZ8+excnJieHDh/PWW29x//332zs8pe5IVolCROS4MebltG8YY8ppslAqa3PmzOHs2bM0bdqU2bNn4+eXbqoVpRxCpoPZxpgfRKRD8qUmIfWMdSIitfIiwHRx6WC2yqfi4uK4evWq9SG5w4cPs2HDBvr160ehQtkZDlTKdmxawiO/0USh8qM//viDgQMHUrlyZdauXasPy6l8x6Z3PRljHjTGlEj++TljzEfGmGo52ZlSBU1ERATPP/88LVu25NChQ5w8eZLz58/bOyylclV2+sOfAjHGGD8slWNPAF/ZNCql8rmkpCQ+++wzPDw8+PLLL3F2dubtt99mz549PPDAA/YOT6lclZ2igAkiIsaYTsA0EfnMGNPH1oEplV+JCO3atWPt2rUAtGnThlmzZuHu7m7nyJSyjez0KKKMMf8GegE/GmOcAH1CSN2zjDE8/PDDVKxYka+//ppff/1Vk4Qq0LJTwuMB4P+AHSLyZ/L4REsRWZgXAaalExcpe/jxxx+Jj4/nqaeeAix3ON24cYMyZcrYOTKlssemg9kicg5YDJQ2xnQAYu2VJACqXywHr22w1+7VPebUqVN06dKFDh060K9fPy5ftjw+5OzsrElC3TOyc9dTN2A78AzQDdhmjOlq68CUsqeEhASmTp2Kp6cn//3vfylRogSjR4+mVKlS9g5NqTyXncHsN4GAW1OfGmNcgbXAclsGppS9bN++nQEDBrB7924AOnfuzLRp03Bzc7NzZErZR3YSRaE082NfInuD4DYRUTIaetW31+5VAZeUlETfvn05cOAA1apVY8aMGTz55JP2Dkspu8rOYPZkwBfLvNkA3YE9Oh+FKihEhLi4OIoVKwbAhg0b+Omnnxg7diwlSpSwc3RK5Q6bl/AwxjwNPISl3tNGEVmRk53lBk0UKjeFhYUxePBg3Nzc+Oyzz+wdjlI2Y5NEYYxxB6YAtYG9wAgROZ3jKHOJJgqVG+Li4pg4cSLvv/8+cXFxlCtXjtDQUFxcXOwdmlI2YavbYz8HfgC6AMHA9JzsQKn85vfff8fX15e33nqLuLg4+vTpw6FDhzRJKJWJrAazS4rIvOSfDxtjduVFQErZSmJiIn379uWrryylyurVq8fs2bNp2bKlfQNTKp/LKlEUM8Y05J95KO5LuSwimjiUQ3FycqJw4cIUK1aM//znP4wYMQJnZ2d7h6VUvpfVGMX6LD4nItLKNiFlrX6JB+RA02mwrrs9dq8czN69e4mNjSUgIACAS5cucfXqVWrXrm3nyJTKW3czRpFpj0JEHs15SLZT/GZR2BNh7zBUPnf9+nXGjRvH1KlTcXd3JyQkhKJFi+Li4qJjEUrdoew8cKeUQ1m1ahVDhw4lPDwcYwxt2rQhPj6eokWL2js0pRySTZ+wNsYEGWMOG2PCjDFvZNEuwBiTqDWk1N0IDw/nqaeeolOnToSHh+Pv78/27duZPn26Pjin1F2wWY8ied6KmUBb4BSwwxizSkQOZNBuIvBLdrZ7sMo5+L5bboerHFxiYiItW7bk2LFjlCxZknfffZfBgwdTuLB2mpW6W9mpHmuS58oem7xczRgTmI1tBwJhInJURG4CS4BOGbQbCnwHXMjgvXRinOPBr0J2mqp7wK2bMZycnBg3bhxdu3bl4MGDDBs2TJOEUrkkO5eeZgHNgJ7Jy1FYegq3UwU4mWL5VPI6K2NMFaAzMDurDRlj+htjdhpjdmZjv+oecOXKFQYOHMj7779vXderVy+WLVtGlSpVsvikUupOZSdRNBGRl4FYABG5AmRnVNBksC7tvbgfA6NEJDGrDYnIXBFpnNNbu1TBISIsXrwYDw8P5syZw8SJE4mMjAQsU5QqpXJfdvrm8cnjCALW+SiSsvG5U0DKAv5VgTNp2jQGliT/By8PPGGMSRCRldnYvrrHhIaGMnjwYNatWwfAww8/zKeffkrp0qXtHJlSBVt2ehSfACuACsaY94C/gPez/ggAOwB3Y0xNY0xRoAewKmUDEakpIjVEpAaWiZAGa5JQaSUkJDBu3Dh8fHxYt24dLi4ufP755/zxxx94eXnZOzylCrzb9ihEZLExJhhojeVy0lMicjAbn0swxgzBcjeTE/C5iOw3xgxMfj/LcYnMFI8rAiEXdED7HuLk5MSff/7JzZs3eeGFF5g4cSLly5e3d1hK3TOyM3FRtYzWi0i4TSK6jcZFqsnOMiMhYog9dq/yyPnz54mNjaV69eoAHDlyhLNnz/LII4/YOTKlHJOtyozf8iOWcuM/AuuAo8BPOdmZUreTlJTE7NmzqVevHi+++KL19ld3d3dNEkrZSXYuPfmkXDbG+AMDbBaRumft3r2bgQMHsm3bNgCKFi1KdHQ0JUuWtHNkSt3b7riER3J58QAbxJItMUVvgq+rvXavbCAqKopXX32VRo0asW3bNipXrsyyZcv48ccfNUkolQ9kZ4zi1RSLhQB/wEVE2tkysEzj0alQC5SbN2/i5eVFWFgYhQoVYujQoYwfP55SpUrZOzSlChSblBlPIeUpXQKWsYrvcrIzpdIqWrQovS//36UAACAASURBVHr1YvXq1cyePZtGjRrZOySlVBpZ9iiSH7T7QERez7uQsqY9CscWHx/P1KlTqVatGj169AAsvQonJyecnJzsHJ1SBZdNehTGmMLJz0L45zw0pf6xadMmBg4cyL59+3B1daVDhw7cf//9Ok+EUvlcVpeetmMZj9htjFkFLAOu33pTRP5r49hUAXH58mVGjRrF/PnzAahVqxazZs3i/vvvt3NkSqnsyM4YRTngEtAKS70nk/ynJgqVJRHhq6++4rXXXuPixYsUKVKEUaNGMXr0aO677z57h6eUyqZMxyiMMaeAj/gnMaQszSki8pHtw0uverlycuL55fBRK3vs/p4THx/PqVOniI2NvePPighnz54lPj4eZ2dnXFxcKFKkiA2iVErdUqxYMapWrZru/5qt7npyAu4ne+XC84xr1P3w1QFNFHnk1KlTlCxZkho1amSrjHdSUhJJSUnWSYPc3NyIi4vDxcVFy4ArZWMiwqVLlzh16hQ1a9bMte1mlSjOisj4XNuTckixsbHZThKRkZGEh4dbEwtAyZIl9aE5pfKIMQYXFxciIiJydbtZJQo9/VPA7ScEunnzJidPnuTKlSsAFCpUiMTERL3dVSk7sEXPPatE0TrX95YLTpS/DBNb2jsMhaWbGxERwenTp0lMTKRQoUJUrlyZChUqUKjQHVeHUUrlU5n+bxaRy3kZSHZdLHUdenvbO4x7XlJSEocOHSI8PJzExERKly6Nl5cXDzzwQK4nCScnJxo0aIC3tzdPPvkkV69etb63f/9+WrVqRd26dXF3d+edd94h5Q0aP/30E40bN8bT0xMPDw9GjBiRq7HZUs+ePfH19WXq1KnZam+r241FhGHDhlGnTh18fX3ZtWtXpu1atWrFtWvXbBJHbggKCqJMmTJ06NAh0zZxcXF0796dOnXq0KRJE44fP25978svv8Td3R13d3e+/PJL6/oePXpw5MgRW4ZuXyLiUC/qIirvHDhwINP3jh07JiEhIXL58mVJSkqyWQwlSpSw/ty7d2959913RUQkJiZGatWqJb/88ouIiFy/fl2CgoJkxowZIiKyd+9eqVWrlhw8eFBEROLj42XmzJm5Glt8fHyubu+Ws2fPSrVq1e7oMym/p9z0448/SlBQkCQlJcmWLVskMDAww3Y//PCDvPLKK3e07YSEhNwIMdvWrl0rq1atkvbt22faZubMmTJgwAAREfnmm2+kW7duIiJy6dIlqVmzply6dEkuX74sNWvWlMuXL4uIyIYNG+Sll16y/QFkU0b/b4GdktPfuzn9oL1emijy1q1/cElJSYLlbrdcf91Oyl+An376qQwaNEhERObPny+9evVK1TYsLEyqVq0qIiK9evWSzz777Lbbj4qKkueff168vb3Fx8dHli9fnm6/y5Ytkz59+oiISJ8+fWT48OHSsmVLeeWVV6R69epy5coVa9vatWvLuXPn5MKFC/L0009L48aNpXHjxvLXX3+l2/eNGzes+27QoIH8/vvvIiLi4+MjxYoVEz8/P9m4cWOqz5w7d06eeuop8fX1FV9fX9m0aVOqeKOioqRVq1bSsGFD8fb2lpUrV4qISHR0tDzxxBPi6+srXl5esmTJEhERGTVqlHh6eoqPj4+89tpr6WLs37+/fP3119blunXrypkzZ9K169mzp6xfv9663KlTJ/H395f69evLnDlzrOtLlCghY8aMkcDAQPnzzz/lq6++koCAAPHz85P+/ftbk8fAgQOlUaNGUr9+fRk7dmy6/eXU+vXrs0wUjz32mGzevFlELCcCLi4ukpSUJF9//bX079/f2i7l95KYmCg1atSw2YnDncrtRJGdB+7UPS4uLo7wcLtMaJhKYmIi69at48UXXwQsl53SFhGsXbs20dHRXLt2jX379vHaa6/ddrvvvPMOpUuXZu/evQDWQfmshIaGsnbtWpycnEhKSmLFihX07duXbdu2UaNGDSpWrMj//d//MXz4cB566CHCw8Np164dBw+mnkV45syZAOzdu5dDhw7x2GOPERoayqpVq+jQoQO7d+9Ot+9hw4bRokULVqxYQWJiItHR0aneL1asGCtWrKBUqVJcvHiRpk2b0rFjR37++WcqV67Mjz/+CFjuUrt8+TIrVqzg0KFDGGNSXda75fTp07i5uVmXq1atyunTp6lUqVKqdps2bWLOnDnW5c8//5xy5cpx48YNAgIC6NKlCy4uLly/fh1vb2/Gjx/PwYMHmThxIps2baJIkSIMHjyYxYsX07t3b9577z3KlStHYmIirVu3Zs+ePfj6+qba5+TJk1m8eHG6mB955BE++eSTDP/ubifl8RYuXJjSpUtz6dKlTL8HsNzAUadOHUJCQgpkYUtNFCpTN2/eJDIykv3795OUlMSuXbuoUqUKrq6uefpMxI0bN2jQoAHHjx+nUaNGtG3bFrD0hjOL407iW7t2LUuWLLEuly1b9rafeeaZZ6x3dXXv3p3x48fTt29flixZQvfu3a3bPXDggPUz165dIyoqKtXtwn/99RdDhw4FwMPDg+rVqxMaGpplmfXff/+dhQsXApbxm9KlS6d6X0QYPXo0GzdupFChQpw+fZrz58/j4+PDiBEjGDVqFB06dODhhx8mISGBYsWK8dJLL9G+ffsMr91LBg/lZvT9Xr58OdWxffLJJ6xYsQKAkydPcuTIEVxcXHBycqJLly4ArFu3juDgYAICLFPc3LhxgwoVKgDw7bffMnfuXBISEjh79iwHDhxIlyhef/11Xn89d2uWZna8t/seKlSowJkzZwpkotBbU1SG/vzzTxo2bMjVq1dJSkqiXLlyeHt7U6FChTx/cO6+++5j9+7dnDhxgps3b1rPwr28vNi5c2eqtkePHuX++++nZMmSeHl5ERwcfNvtZ5ZwUq5L+2R6iRIlrD83a9aMsLAwIiIiWLlyJU8//TRgGfDfsmULu3fvZvfu3Zw+fTrdMyUZ/fK5W4sXLyYiIoLg4GB2795NxYoViY2NpW7dugQHB+Pj48O///1vxo8fT+HChdm+fTtdunRh5cqVBAUFpdte1apVOXnypHX51KlTVK5cOV27woULk5SUBMCGDRtYu3YtW7ZsISQkhIYNG1q/w2LFilmTrIjQp08f63d0+PBhxo0bx7Fjx5gyZQrr1q1jz549tG/fPsPqAJMnT6ZBgwbpXsOGDcvx95fyeBMSEoiMjKRcuXK3/R5iY2MLbGkah0sUjY66gesMe4dRoN24cYOuXbty4MABChcuTN26dalVq5bdy2+ULl2aTz75hClTphAfH8+zzz7LX3/9xdq1a61xDxs2jJEjRwKWs83333+f0NBQwPKL+6OP0leeeeyxx5gx459/U7cuPVWsWJGDBw9aLy1lxhhD586defXVV/H09MTFxSXD7WZ0GemRRx6xXjoJDQ0lPDycevXqZfk9tG7dmk8//RSwXI5Le5dRZGQkFSpUoEiRIqxfv54TJ04AcObMGYoXL85zzz3HiBEj2LVrF9HR0URGRvLEE0/w8ccfZxhjx44dWbhwISLC1q1bKV26dLrLTgD16tXj6NGj1hjKli1L8eLFOXToEFu3bs30WJYvX86FCxcAS6/kxIkTXLt2jRIlSlC6dGnOnz/PTz/9lOHnX3/9dWuSSfnK6WWnW8d7646m5cuX06pVK4wxtGvXjl9//ZUrV65w5coVfv31V9q1+2f+ttDQULy8vHK833wtp4Mb9no1KuwmUn76nY7tqNtISkpKNRC3aNEiGTt2rOzfv9+OUVmkvZunQ4cOsnDhQhER2bNnj7Ro0ULq1q0rtWvXlnHjxqW6A2v16tXi7+8vHh4e4unpKSNGjEi3/aioKOndu7d4eXmJr6+vfPfddyJiGcCuVauWtGjRQl5++eVUg9nLli1LtY0dO3YIIAsWLLCui4iIkG7duomPj494enpa76RJ6caNG9KnT590g9nHjh0TLy+vDL+Pc+fOSceOHcXb21v8/PysA6+3vqeIiAhp2rSpNGrUSF588UXx8PCQY8eOyc8//yw+Pj7i5+cnjRs3lh07dsiZM2ckICBAfHx8xNvbO1X8tyQlJcngwYOlVq1a4u3tLTt27MgwrvHjx8u8efNERCQ2NlaCgoLEx8dHunbtKi1atLAOdKf9+1yyZIn4+fmJj4+P+Pv7y5YtW6zfs4eHhzzxxBPSuXNn+eKLLzLc75146KGHpHz58lKsWDGpUqWK/PzzzyIiMmbMGPn+++9FxPJ30rVrV6ldu7YEBATI33//bf38Z599JrVr15batWvL559/bl1/7tw5CQgIuOv4cktuD2bfdirU/KZxkWqys8xIiBhi71AKjAMHDjBw4EDatm3LmDFjUr138OBBPD097RSZciRnz56ld+/e/Pbbb/YOJc9NnTqVUqVKWW+0sLeM/t/eTVFAh7v0pHJPTEwMo0ePxs/Pjz///JP58+cTFxdn77CUg6pUqRL9+vXL1w/c2UqZMmXo06ePvcOwGYdLFMG1TmpvIhf89NNPeHt7M2HCBBISEhgwYAC7d+/G2dnZ3qEpB9atW7cs79gqqPr27WutmFwQFdwjUxm6fv06zz//PMuXLwfA19eX2bNn06xZMztHppTKrxyuR6HuTvHixbl8+TIlSpRgypQpBAcHa5JQSmVJexT3gJ07d1KmTBnq1KmDMYb58+fj5OREtWrV7B2aUsoBaI+iAIuMjGTo0KEEBgYycOBA68NdNWvW1CShlMo2TRQFkIiwdOlSPDw8mDFjBoUKFcLf35+EhAR7h5YjWmbcvmXGDx06RLNmzXB2dmbKlCmZthPRMuMFVk4fwLDXq/wDJUS+3JuDR1DuDWFhYdKuXTtrZdZmzZpJSEhIjreXVZnxvKJlxrPHVmXGz58/L9u3b5fRo0fL5MmTM22nZca1zHi+eemT2Zm7du2alClTRgApU6aMzJkzRxITE+9qmyn/wbEWm7xuR8uM27fM+C1vvfVWlolCy4xrmfEcMcYEAdMAJ2C+iHyQ5v1ngVHJi9HAIBEJsWVMBVnJkiUZPnw4YWFhTJkyxVqFs6DQMuMWeV1mPLu0zLiWGb9jxhgnYCbQFjgF7DDGrBKRAymaHQNaiMgVY8zjwFygia1iKmgiIiJ4/fXXad26Nb169QJgzJgxNqvuKq3tU+5Fy4ynltdlxrNLy4xrmfGcCATCROSoiNwElgCdUjYQkc0icuv0bStQ9XYbjSgZDb3q53qwjiQpKYn58+dTr149vvzyS958803i4+OBO/sF6Si0zPidye0y49mlZca1zHhOVAFOplg+lbwuMy8CGdYSNsb0N8bsNMbsDHe9Ah+1ysUwHcu+fft45JFH6NevH1euXKFNmzasW7fO7iXA84KWGbfI6zLj2aVlxrXM+J2PksMzWMYlbi33AqZn0vZR4CDgctvt3qNzZsfExMjIkSOlcOHCAkjFihXl66+/TlVS2xby211PIlpmPK/LjJ89e1aqVKkiJUuWlNKlS0uVKlUkMjIyXTstM65lxu+YMaYZME5E2iUv/zs5MU1I084XWAE8LiKht91uPSNy2LFKo+eGuLg4GjRowOHDhxk0aBDvvfceZcqUsfl+tcy4yi4tM15wy4zb8q6nHYC7MaYmcBroAfxfygbGmGrAf4Fe2UkS95pTp05RvHhxypUrh7OzMwsWLACgSRMd71f5T8oy4/daBdkyZcpYbygpiGw2RiEiCcAQ4Bcsl5W+FZH9xpiBxpiByc3GAi7ALGPMbmPMzkw2d09JSEhg6tSpeHp6prqjo0mTJpokVL6mZcYLJpsemYisAdakWTc7xc8vAS/ZMgZHs23bNgYMGEBIiOVxksjISBISEgr0P0KlVP7mcLWePE9VhNZL7R1Grrt69SqDBw+mWbNmhISEUL16dVavXs3y5cs1SSil7MrhfgMVv1kU9kTYO4xcdeXKFerXr8+5c+coXLgwr732GmPGjEl1r75SStmLwyWKgqhs2bI8/vjjhIaG8umnn+Lj42PvkJRSysrhLj0VBHFxcYwfP54//vjDum7GjBls3LhRk0QOrVq1ig8++OD2DQu4BQsW4OrqSoMGDfDw8EhXonzu3Ll4eHjg4eFBYGAgf/31l/W9+Ph43njjDdzd3fH29iYwMDDTB93s6ZVXXmHjxo32DiNTb775Jm5ubrct+z5hwgTq1KlDvXr1+OWXX6zrbz09X6dOHYYNG2Z9en/GjBl88cUXNo09Uzl9AMNer+LVi4jsPn9HD5/kJ+vWrZO6desKIJ6ennleZvlOpXtwp/z01K/MfLk3dbvh62wb6B1ISkq666q6d8OWFUa/+OILefnll0VE5OLFi+Li4iLh4eEi8s/DhxERESIiEhwcLG5ubnL27FkRsVSR7d27t8TGxoqI5SGypUuX5mp8d/vv/dKlS9KkSZM7+kxeV3TdsmWLnDlzJsuy7/v37xdfX1+JjY2Vo0ePSq1atazfTUBAgGzevFmSkpIkKChI1qxZIyKWMvoNGjTIVgy5/cCdw/UoYpzjwc/xqqJeuHCBXr160bp1a0JDQ/Hw8GDWrFnWmjcqY8ePH8fDw4OXXnoJb29vnn32WdauXcuDDz6Iu7s727dvByxn0kOGDAHg/PnzdO7cGT8/P/z8/Ni8eTPHjx/H09OTwYMH4+/vz8mTJ3n99dfx9vbGx8eHpUszvkFi+/btNG/enIYNG9K8eXMOHz4MWG5V3r9/v7Vdy5YtCQ4O5vr167zwwgsEBATQsGFDvv/+e2t8zzzzDE8++SSPPfYY0dHRtG7dGn9/f3x8fKztwFLN1sPDg7Zt29KzZ0/rZEF///03QUFBNGrUiIcffphDhw5l+d25uLhQp04dzp49C8DEiROZPHky5cuXB8Df358+ffowc+ZMYmJimDdvHtOnT8fZ2RmwlDDp1q1buu3u2LGD5s2b4+fnR2BgIFFRUam+f4AOHTqwYcMGwDKh0tixY2nSpAnvv/9+qm1u2LCBJ598EoBff/2VZs2a4e/vzzPPPJOuKi5YSmqkrEc1fvx4AgIC8Pb2pn///taz75YtWzJ69GhatGjBtGnTCA4OpkWLFjRq1Ih27dpZv5N58+YREBCAn58fXbp0ISYmJsvvNDuaNm1KpUqVsmzz/fff06NHD5ydnalZsyZ16tRh+/btnD17lmvXrtGsWTOMMfTu3ZuVK1cClvnua9SoYf03n6dymmHs9XK0Eh6JiYkyZ84c6zwRxYoVk3fffVfi4uLsHVq22LtHcezYMXFycpI9e/ZIYmKi+Pv7S9++fSUpKUlWrlwpnTp1EpHUZ9LdunWTqVOniojlDPbq1aty7NgxMcZYy0MsX75c2rRpIwkJCXLu3Dlxc3OTM2fOpNt/ZGSk9Yz0t99+k6efflpERD766CPrHAlnzpwRd3d3ERH597//LV999ZWIiFy5ckXc3d0lOjpavvjiC6lSpYpcunRJRCxnubfKYEREREjt2rUlKSlJduzYIX5+fhITEyPXrl2TOnXqWOeAaNWqlYSGhoqIyNatW+XRRx9NF2/K7+HEiRPi5+cnN27cEBGRsmXLytWrV1O1X7lypXTu3FlCQkKydbYaFxcnNWvWlO3bt6f6flLuV0Skffv21pIdgLVnEh8fL25ubhIdHS0iljknvvrqK4mIiJCHH37Yuv6DDz6Qt99+O93+e/fuLatWrbIu3/o+RUSee+4563stWrSwzlty8+ZNadasmVy4cEFELCVD+vbtKyKWXtctb775pnzyySfp9vn777+Ln59fulezZs2y/K6y6lG8/PLL1n8nIiIvvPCCLFu2THbs2CGtW7e2rt+4cWOquTPeffddmTJlSpb7FXGw+SiU5TmIN998k6tXr9KuXTtmzpxJ7dq17R2WQ6lZs6Z17MbLy4vWrVtjjMHHxyfVNJW3ZFSG+8qVK1SvXp2mTZsClvLePXv2xMnJiYoVK9KiRQt27NhBx44dU20rMjKSPn36cOTIEYwx1iq93bp1o23btrz99tt8++23PPPMM4DlrHjVqlXWXkBsbCzh4eEAtG3blnLlygGZlwL/66+/6NSpk7UK6a2z7ejoaDZv3mzdD1jGujKydOlS1q9fz+HDh5k3bx7FihXL9LsVybxUe0YOHz5MpUqVrGXBs/NwXcqy4oULFyYoKIjVq1fTtWtXfvzxRyZNmsQff/zBgQMHePDBBwG4efMmzZo1S7ets2fP4urqal1ev349kyZNIiYmhsuXL+Pl5WX9zm6Vez98+DD79u2zlqdPTEy0nvHv27eP//znP1y9epXo6OhURf5uefTRR++qWGJGJIPSSdktZX67nqQtaKKwgevXr1O4cGGcnZ0pW7Yss2fPJjExkWeeecbxy4BHDLl9G4De3pZXLrh1KQQsE8TcWi5UqNAdzQOe8nbjjP5DgmUioXnz5gGwZs0axowZw6OPPsqKFSs4fvw4LVu2BKBKlSq4uLiwZ88eli5dap2wR0T47rvv0lWA3bZtW6r9pywFXqRIEWrUqEFsbGymcSUlJVGmTJls/cLq3r07M2bMYMuWLbRv357HH3+cBx54gPr16xMcHEyrVv9UX961axf169enTp06hIeHp5svI63MEkvKEuOQuix7yrLit+KbOXMm5cqVIyAggJIlSyIitG3blm+++SbLY7vvvvus246NjWXw4MHs3LkTNzc3xo0bl2q/t75vEcHLy4stW7ak297zzz/PypUr8fPzY8GCBdbLZSmtX7+e4cOHp1tfvHhxNm/enGW8mcmsZHnVqlU5depUuvW32KuUucONUeR3q1aton79+kyaNMm6rkuXLnTr1s3xk4SDuF0ZbrCU9166dCmJiYlERESwceNGAgMDefnll62lqitXrkxkZCRVqliq49+qtXVLjx49mDRpEpGRkdYeT7t27Zg+fbr1F/7//ve/DGPMrBT4Qw89xOrVq4mNjSU6Oto6G12pUqWoWbMmy5YtAyy//G49vZ+ZZs2a0atXL6ZNmwbAyJEjGTVqFJcuXQIsZc8XLFjA4MGDKV68OC+++CLDhg3j5s2bgOXsfdGiRam26eHhwZkzZ9ixYwcAUVFRJCQkUKNGDXbv3k1SUhInT57M8jp6y5Yt2bVrF/PmzbOe9Tdt2pRNmzYRFhYGQExMjLU8fEqenp7WNreSQvny5YmOjmb58uUZ7q9evXpERERYE0V8fLx1fCkqKopKlSoRHx+f4Ux58E+PIu0rp0kCLKXMlyxZQlxcHMeOHePIkSMEBgZSqVIlSpYsydatWxERFi5cSKdO/0zjExoaird37pyA3QmHSxTF44pAyAV7h5FOeHg4Tz31FJ06dSI8PJxffvkl1RmWyjvTpk1j/fr1+Pj40KhRo1SDzrd07twZX19f/Pz8aNWqFZMmTeKBBx5I127kyJH8+9//5sEHHyQxMTHVe127dmXJkiWpBmfHjBlDfHw8vr6+eHt7M2bMmAxjfPbZZ9m5cyeNGzdm8eLFeHh4ABAQEEDHjh3x8/Pj6aefpnHjxtYZ7BYvXsxnn32Gn58fXl5eqQbAMzNq1Ci++OILoqKi6NixIy+88ALNmzfHw8ODfv36sWjRIutlmHfffRdXV1fq16+Pt7c3Tz31VKrLPABFixZl6dKlDB06FD8/P9q2bUtsbCwPPvig9RLhiBEj8Pf3zzQmJycnOnTowE8//WSdUc/V1ZUFCxbQs2dPfH19adq0aYaXWNq3b2896y9Tpgz9+vXDx8eHp556yno5LK2iRYuyfPlyRo0ahZ+fHw0aNLD+kn/nnXdo0qQJbdu2tf4d3K2RI0dStWpVYmJiqFq1KuPGjQMsJ5Fjx44FLJdQu3XrRv369QkKCmLmzJnWXtenn37KSy+9RJ06dahduzaPP/64ddubNm2iTZs2uRLnHcnp4Ia9Xo0Ku2U9iJrHbt68KZMnT5bixYsLICVLlpRp06bl+9tesys/zEdxr4mKihIRy+2QjRo1kuDgYDtHlL88+OCDcuXKFXuHked27dolzz33XLba6mB2PnLx4kXrpO9gmUd56tSp1ksVSuVE//79OXDgALGxsfTp0yfLs/N70Ycffkh4eHiezMeSn1y8eJF33nnHLvvWRHEXXFxcKF++PDVr1mTGjBk88cQT9g5JFQBff/21vUPI1+7VUvu37tqyB4dLFDFFb4Kv6+0b2oCIsHjxYgIDA6lbty7GGBYtWkTp0qUpXry4XWJSSilbc7jB7INVz8O67nm+38OHD9OmTRt69erF4MGDrXe1VKpUSZOEUqpAc7hEkddiY2N566238PX15ffff8fFxYXnnnvO3mEppVSecbhLT3lp7dq1DBo0yHrf9gsvvMCkSZNwcXGxc2RKKZV3tEeRifPnz9OhQwfCwsKoX78+Gzdu5LPPPtMkoRzG8ePHue+++2jQoAH169end+/e1hIkYCljEhgYaC07Pnfu3FSfX7hwId7e3nh5eVG/fn1rWZL8ZOXKlYwfP97eYWRq2bJleHl5UahQIXbu3Jlpu59//pl69epRp06dVOXyL1++TNu2bXF3d6dt27ZcuXIFgL179/L888/bOvx/5PS+Wnu9bFkUMDExUZKSkqzLEydOlAkTJjhMAT9bSHs/NoxL9crMnDk7U7Xr129Vpm3tzZ7PvNiy5PmxY8fEy8tLRCzH+Oijj8qiRYtEROTs2bPi5uZmfUYjIiJC/P395YcffhARkTVr1kjDhg3l9OnTIiJy48YNmTt3bq7Glxvlv5s1a2Ytm55X+7wTBw4ckEOHDkmLFi1kx44dGbZJSEiQWrVqyd9//y1xcXHi6+sr+/fvFxGR119/XSZMmCAiIhMmTJCRI0daP9e6dWs5ceJEpvtNi3upzLit7N69m+bNm6cqWTBy5EjeeOMNihYtasfI7m3ZLTOeWTnwxMRERowYgY+PD76+vkyfPh2AGjVqMH78eB566CGWLVvGN998MQ5tNQAAEdxJREFUg4+PD97e3owaNSrDWDIrDT5q1ChmzZplbTdu3Dg+/PBDACZPnkxAQAC+vr689dZb1mNKW/J80KBBNG7cGC8vL2s7sNSb8vDw4KGHHmLYsGHWJ5kzK2eeGScnJwIDAzl9+jRgqWn1/PPPW5/RKF++PJMmTbKezU6YMIEpU6ZY6wwVK1aMfv36pdtuZiXdU5aZmDJlivXp5JTlv9977z1q1KhhrWAQExODm5sb8fHx2SqpHhoairOzs7Vs+urVq2nSpAkNGzakTZs2nD9/3vr30b9/fx577DF69+5NREQEXbp0ISAggICAADZt2gRk/m/obnh6eqar+5XW9u3bqVOnDrVq1aJo0aL06NHD+vf5/fff06dPHwD69OljLTkOloKRS5YsuesYsyWnGcZer2ply+bqJDjXrl2T4cOHS6FChQSQBg0apOpV3Ovs3aPIbpnxzMqBz5o1S55++mnre7fKUlevXl0mTpwoIiKnT58WNzc3uXDhgsTHx8ujjz4qK1asSBdLZqXBd+3aJY888oi1naenp5w4cUJ++eUX6devn7XX0L59e/njjz/SlTxPGVdCQoK0aNFCQkJC5MaNG1K1alU5evSoiIj06NHDWnI6s3Lmab+7Wz2KGzduSMuWLSUkJERE5P/bO//gKMo0j38eIAoLGFIiVCSbFT3cQEpCgBP1RGGXiyxihDKK6LGFiIiygZPb44duLZZ3t2QXpG5jShA4irisREVlXe9cDacQVNCAoCKgIHhuJBiTrCQSFSHP/dE9nUmYmTQhM5mE51PVNd3Tb7/v08/09NPvj/6+OnHiRN24cWOj9F999ZUmJSWpamhJ8lCEk3QPlKuqumTJEl20aJGqNpb/VlXNzs7W1157TVUd+e+7775bVf1Jqq9Zs0bnzp3rbVdXV3v/3VWrVnn7Fi1apEOHDtW6ujpVVZ08ebJu3bpVVR0p9rS0NFUNfw0FU1NTE1JyPCMjw6sFhCJSjeLZZ5/1zltV9cknn/Qk2xMTExul7dWrl7f+xhtv6Pjx40Pmec6/mX1RbQ/4w15Y9pPmE0dAVdm4cSOzZ8+mrKyMTp06MWfOHB555BET74sz/MiMh5MD37RpEzNnzqRLF+dSD8h8Q4MMdWlpKaNGjfJ0je68805KSkqYMGFCIztUQ0uDZ2ZmUlFRwZEjR/jyyy9JSkoiNTWV/Px8Xn31VTIzMwGnRnLgwAFSU1MbSZ4DPPPMM6xcuZKTJ09SXl7O3r17qa+v59JLL6V///4ATJ482etHCCdnPnDgwEY2f/LJJwwZMoQDBw6Qk5PD4MGDvXMJdZ2f6bUfTtI9EgG/B9affvppRo8eTVFREffff79vSfWmkuNlZWVMmjSJ8vJyTpw44fkNHBG+gOrqpk2b2Lt3r7evpqaG2trasNdQMD179oyZ5Hhz9OnThyNHjrSqLeFod4GiNaisrOSuu+7ipZdeAmD48OE88cQTJpXgA9VFzScCZswYxowZw1qlTD8y4+HkwMPdEKGxDHUo3n77be69917AmUmturo6pDQ4OAKBGzZs4OjRo9x+++1evgsXLvTyCPDpp582khw/fPgwS5cupbS0lKSkJKZOnRpRcjyQdyg586Zcdtll7N69m/LyckaNGsWLL75IdnY26enp7Nixo9H8Gzt37mTQoEGAE5CbSpL7JZLkODSWe8/OzmbhwoVUV1d75R0/ftyXpHq3bt04duyYt52bm8vcuXPJzs5m8+bNXnNX0zLr6+vZtm3baXLdubm5Ia+hYGpraxk5cmRIe5566inPf2dCOMlxcGYZLC8vJzk5mfLycvr0aZjdM5aS4+dkH0XPnj05ePAgF1xwAQUFBWzfvt2CRDsnnBx4VlYWK1as8AJKdXX1aceOGDGCLVu2UFlZyalTp1i/fj3XX389I0aM8CSls7Ozw0qDgyM5XlRUxIYNG8jJyQEcyfE1a9Z4U3p+/vnnVFScrnxcU1ND9+7dSUxM5IsvvuDll18GHEnvQ4cOebWm4Ola/cqZB0hOTiYvL4/FixcDMGvWLNauXevdjKuqqpg/fz7z5s0DYOHChcybN4+jR48CzhN9fn7+afmGknTv27cvFRUVVFVV8d1333kPZKHo0aMHV155JXPmzGH8+PF07tzZt6R6sOQ4NL4GCgsLw5aZlZVFQUGBtx3wQSRJ+QCBGkWopSVBAhzF4AMHDnD48GFOnDhBUVGRF8Czs7O9cyksLGwzyfF2Fyj+r3c1PDrqjI978803PR3+888/n6KiIvbv38+sWbNs3uoOQDg58OnTp5OamupJiofSUUpOTmbx4sWMHj2ajIwMhg4d2ugPGSCcNDg4T+C1tbX069fPk+3Oysrijjvu4Oqrr+aKK64gJyeH2tra0/LNyMggMzOT9PR0pk2b5s3y1q1bNx5//HHGjh3LtddeS9++fT3Jcb9y5sFMmDCBuro6tm7dSnJyMuvWreOee+4hLS2Na665hmnTpnmzw40bN45Zs2YxZswY0tPTGTZsWMhJokJJuickJHhzZI8fP75Z+e5Jkyaxbt26Rk1SfiTVr7vuOnbt2uUFy4cffphbb72VkSNHeh3cocjPz2fHjh0MHjyYQYMGsWLFCiCypHxLeeGFF0hJSfEmkQrMoHfkyBFPG65Lly4UFBRwww03MHDgQG677TbS09MBWLBgAcXFxQwYMIDi4mIWLFjg5f36669z4403toqdzdLSzo22Ws50eGxlZaVOnz5dgUYdRoY/TGa8bQlIjtfX1+t9992ny5Yta2OL4ovZs2drcXFxW5sRc7799lsdMWJE2OG+NjzWJ6pKYWEhaWlprF69moSEBC6++OKI7b6GEW+sWrWKIUOGkJ6ezrFjx07r7zjXefDBB6mrq2trM2LOZ599Rl5enjdII9pIe7txyo9F9aPINu/fv5+ZM2eyZcsWwBm7vXz58labwepcYt++faeNpDEMI74J9b8VkZ2qOrwl+XW4UU9lZWVkZGRw4sQJevfuzaOPPsqUKVNsyOtZoBFGDhmGEV9E4+G/wwWKlJQUpkyZQqdOncjLy2s0bt44c7p27UpVVRUXXnihBQvDiHNUlaqqKrp27dqq+bb7pqfy8nIeeOABZs6c6Y17rq+vp1OnDtv9ElO+//57ysrKThsLbxhGfNK1a1dSUlJISEho9P051fQ07NAP4aICTh29j+XLl/PQQw9RU1PDwYMHKS0tRUQsSLQiCQkJjd5wNQzj3COqd1QRGSsiH4nIQRFZEGK/iEi+u/99EfH11tu7J//KVVddRW5uLjU1Ndx0000899xz1jRiGIYRBaLW9CQinYGPgX8EyoBSYLKq7g1KMw7IBcYBI4Dfq2rEmdP7duqplXqcepSUlBQee+wxbr75ZgsShmEYETibpqdo1iiuBA6q6iFVPQEUAU1fd70ZeNJ9H2Q70EtEkiNlWq11CMLcuXPZt28fEyZMsCBhGIYRRaLZR9EP+GvQdhlOraG5NP2A8uBEIjIDmOFufgfsWbZsGcuWLWtVg9shvYHKtjYiTjBfNGC+aMB80UBkBckIRDNQhHrMb9rO5ScNqroSWAkgIjtaWn3qaJgvGjBfNGC+aMB80YCIhJ+LtRmi2fRUBvwwaDsFaCqe7ieNYRiG0YZEM1CUAgNEpL+InAfcDrzYJM2LwM/d0U9XAcdUtbxpRoZhGEbbEbWmJ1U9KSK/AF4BOgNrVPVDEZnp7l8B/A/OiKeDQB1wl4+sV0bJ5PaI+aIB80UD5osGzBcNtNgX7e7NbMMwDCO22CvMhmEYRkQsUBiGYRgRidtAES35j/aID1/c6frgfRF5S0Qy2sLOWNCcL4LS/b2InBKRnFjaF0v8+EJERonIbhH5UES2xNrGWOHjP5IoIn8WkfdcX/jpD213iMgaEakQkT1h9rfsvtnSqfGiueB0fn8CXAqcB7wHDGqSZhzwMs67GFcBb7e13W3oi2uAJHf9Z+eyL4LSvYYzWCKnre1uw+uiF7AXSHW3+7S13W3oiweB37rrFwHVwHltbXsUfHEdMBTYE2Z/i+6b8VqjiIr8RzulWV+o6luq+jd3czvO+ygdET/XBTj6Yc8BFbE0Lsb48cUdwPOq+hmAqnZUf/jxhQI9xdH76YETKE7G1szoo6olOOcWjhbdN+M1UIST9jjTNB2BMz3Pu3GeGDoizfpCRPoBE4EVMbSrLfBzXVwOJInIZhHZKSI/j5l1scWPLwqAgTgv9H4AzFHV+tiYF1e06L4Zr/NRtJr8RwfA93mKyGicQHFtVC1qO/z44j+B+ap6qoOLRfrxRRdgGPBToBuwTUS2q+rH0TYuxvjxxQ3AbuAnwGVAsYhsVdWaaBsXZ7TovhmvgcLkPxrwdZ4iMhhYDfxMVatiZFus8eOL4UCRGyR6A+NE5KSqboyNiTHD73+kUlWPA8dFpATIwJH/70j48cVdQJ46DfUHReQwkAa8ExsT44YW3TfjtenJ5D8aaNYXIpIKPA9M6YBPi8E06wtV7a+ql6jqJcAG4P4OGCTA33/kT8BIEekiIj/AUW/eF2M7Y4EfX3yGU7NCRPriKKkeiqmV8UGL7ptxWaPQ6Ml/tDt8+uLXwIXA4+6T9EntgIqZPn1xTuDHF6q6T0T+ArwP1AOrVTXksMn2jM/r4t+AtSLyAU7zy3xV7XDy4yKyHhgF9BaRMmARkABnd980CQ/DMAwjIvHa9GQYhmHECRYoDMMwjIhYoDAMwzAiYoHCMAzDiIgFCsMwDCMiFiiMuMRVft0dtFwSIe3XrVDeWhE57Jb1rohc3YI8VovIIHf9wSb73jpbG918An7Z46qh9mom/RARGdcaZRvnLjY81ohLRORrVe3R2mkj5LEWeElVN4hIFrBUVQefRX5nbVNz+YpIIfCxqv5HhPRTgeGq+ovWtsU4d7AahdEuEJEeIvK/7tP+ByJymmqsiCSLSEnQE/dI9/ssEdnmHvusiDR3Ay8B/s49dq6b1x4R+Wf3u+4i8t/u3AZ7RGSS+/1mERkuInlAN9eOP7r7vnY/nw5+wndrMreISGcRWSIipeLME3CvD7dswxV0E5ErxZmLZJf7+WP3LeVHgEmuLZNc29e45ewK5UfDOI221k+3xZZQC3AKR8RtN/ACjorABe6+3jhvlgZqxF+7n/8CPOSudwZ6umlLgO7u9/OBX4coby3u3BXArcDbOIJ6HwDdcaSpPwQygVuAVUHHJrqfm3Ge3j2bgtIEbJwIFLrr5+EoeXYDZgC/cr8/H9gB9A9h59dB5/csMNbdvgDo4q6PAZ5z16cCBUHH/wb4J3e9F47uU/e2/r1tie8lLiU8DAP4RlWHBDZEJAH4jYhchyNH0Q/oCxwNOqYUWOOm3aiqu0XkemAQ8KYrb3IezpN4KJaIyK+AL3FUeH8KvKCOqB4i8jwwEvgLsFREfovTXLX1DM7rZSBfRM4HxgIlqvqN29w1WBpm5EsEBgCHmxzfTUR2A5cAO4HioPSFIjIARw00IUz5WUC2iPzS3e4KpNIxNaCMVsIChdFeuBNnZrJhqvq9iHyKc5PzUNUSN5DcCPxBRJYAfwOKVXWyjzL+VVU3BDZEZEyoRKr6sYgMw9HMWSwir6rqI35OQlW/FZHNOLLXk4D1geKAXFV9pZksvlHVISKSCLwEzALycbSMXlfViW7H/+Ywxwtwi6p+5MdewwDrozDaD4lAhRskRgM/appARH7kplkF/BfOlJDbgX8QkUCfww9E5HKfZZYAE9xjuuM0G20VkYuBOlVdByx1y2nK927NJhRFOGJsI3GE7HA/7wscIyKXu2WGRFWPAbOBX7rHJAKfu7unBiWtxWmCC/AKkCtu9UpEMsOVYRgBLFAY7YU/AsNFZAdO7WJ/iDSjgN0isgunH+H3qvolzo1zvYi8jxM40vwUqKrv4vRdvIPTZ7FaVXcBVwDvuE1ADwH/HuLwlcD7gc7sJryKM7fxJnWm7gRnLpG9wLsisgd4gmZq/K4t7+HIav8Op3bzJk7/RYDXgUGBzmycmkeCa9sed9swImLDYw3DMIyIWI3CMAzDiIgFCsMwDCMiFigMwzCMiFigMAzDMCJigcIwDMOIiAUKwzAMIyIWKAzDMIyI/D/D1/spD7pHdwAAAABJRU5ErkJggg==\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "ename": "ExecutableNotFound",
- "evalue": "failed to execute ['dot', '-Tpng'], make sure the Graphviz executables are on your systems' PATH",
- "output_type": "error",
- "traceback": [
- "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
- "\u001b[0;31mFileNotFoundError\u001b[0m Traceback (most recent call last)",
- "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/graphviz/backend.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(cmd, input, capture_output, check, encoding, quiet, **kwargs)\u001b[0m\n\u001b[1;32m 163\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 164\u001b[0;31m \u001b[0mproc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msubprocess\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mPopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcmd\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstartupinfo\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mget_startupinfo\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 165\u001b[0m \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/subprocess.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, args, bufsize, executable, stdin, stdout, stderr, preexec_fn, close_fds, shell, cwd, env, universal_newlines, startupinfo, creationflags, restore_signals, start_new_session, pass_fds, encoding, errors)\u001b[0m\n\u001b[1;32m 728\u001b[0m \u001b[0merrread\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merrwrite\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 729\u001b[0;31m restore_signals, start_new_session)\n\u001b[0m\u001b[1;32m 730\u001b[0m \u001b[0;32mexcept\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/subprocess.py\u001b[0m in \u001b[0;36m_execute_child\u001b[0;34m(self, args, executable, preexec_fn, close_fds, pass_fds, cwd, env, startupinfo, creationflags, shell, p2cread, p2cwrite, c2pread, c2pwrite, errread, errwrite, restore_signals, start_new_session)\u001b[0m\n\u001b[1;32m 1363\u001b[0m \u001b[0merr_msg\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;34m': '\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mrepr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr_filename\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1364\u001b[0;31m \u001b[0;32mraise\u001b[0m \u001b[0mchild_exception_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merrno_num\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merr_msg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0merr_filename\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 1365\u001b[0m \u001b[0;32mraise\u001b[0m \u001b[0mchild_exception_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr_msg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: 'dot': 'dot'",
- "\nDuring handling of the above exception, another exception occurred:\n",
- "\u001b[0;31mExecutableNotFound\u001b[0m Traceback (most recent call last)",
- "\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m\u001b[0m\n\u001b[1;32m 41\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 42\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 43\u001b[0;31m \u001b[0mxgb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot_tree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mxg_clf\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mnum_trees\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 44\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrcParams\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'figure.figsize'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;36m50\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m10\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 45\u001b[0m \u001b[0msave_fig\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"xgtree\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/xgboost/plotting.py\u001b[0m in \u001b[0;36mplot_tree\u001b[0;34m(booster, fmap, num_trees, rankdir, ax, **kwargs)\u001b[0m\n\u001b[1;32m 246\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 247\u001b[0m \u001b[0ms\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mBytesIO\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 248\u001b[0;31m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpipe\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'png'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 249\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mseek\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 250\u001b[0m \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mimage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/graphviz/files.py\u001b[0m in \u001b[0;36mpipe\u001b[0;34m(self, format, renderer, formatter, quiet)\u001b[0m\n\u001b[1;32m 136\u001b[0m out = backend.pipe(self._engine, format, data,\n\u001b[1;32m 137\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mformatter\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mformatter\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 138\u001b[0;31m quiet=quiet)\n\u001b[0m\u001b[1;32m 139\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 140\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/graphviz/backend.py\u001b[0m in \u001b[0;36mpipe\u001b[0;34m(engine, format, data, renderer, formatter, quiet)\u001b[0m\n\u001b[1;32m 242\u001b[0m \"\"\"\n\u001b[1;32m 243\u001b[0m \u001b[0mcmd\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcommand\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mformat\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mformatter\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 244\u001b[0;31m \u001b[0mout\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcmd\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcapture_output\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcheck\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mquiet\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mquiet\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 245\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 246\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/graphviz/backend.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(cmd, input, capture_output, check, encoding, quiet, **kwargs)\u001b[0m\n\u001b[1;32m 165\u001b[0m \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 166\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0merrno\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0merrno\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mENOENT\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 167\u001b[0;31m \u001b[0;32mraise\u001b[0m \u001b[0mExecutableNotFound\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcmd\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 168\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 169\u001b[0m \u001b[0;32mraise\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
- "\u001b[0;31mExecutableNotFound\u001b[0m: failed to execute ['dot', '-Tpng'], make sure the Graphviz executables are on your systems' PATH"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- ""
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
+ "execution_count": 15,
+ "metadata": {
+ "collapsed": false
+ },
+ "outputs": [],
"source": [
"\n",
"import matplotlib.pyplot as plt\n",
@@ -1933,7 +1458,7 @@
"\n",
"y_test = xg_clf.predict(X_test_scaled)\n",
"\n",
- "print(\"Test set accuracy with xgboost and scaled data: {:.2f}\".format(xg_clf.score(X_test_scaled,y_test)))\n",
+ "print(\"Test set accuracy with Random Forests and scaled data: {:.2f}\".format(xg_clf.score(X_test_scaled,y_test)))\n",
"\n",
"import scikitplot as skplt\n",
"y_pred = xg_clf.predict(X_test_scaled)\n",
@@ -1959,34 +1484,9 @@
"save_fig(\"xgparams\")\n",
"plt.show()"
]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": []
}
],
- "metadata": {
- "kernelspec": {
- "display_name": "Python 3",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython",
- "version": 3
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python",
- "pygments_lexer": "ipython3",
- "version": "3.6.8"
- }
- },
+ "metadata": {},
"nbformat": 4,
"nbformat_minor": 4
}
diff --git a/doc/src/week45/week45.do.txt b/doc/src/week45/week45.do.txt
index dbe502765..f29755c32 100644
--- a/doc/src/week45/week45.do.txt
+++ b/doc/src/week45/week45.do.txt
@@ -734,27 +734,9 @@ plt.show()
!ec
-!split
-===== Additive boosting for Regression =====
-
-Here we present "Drucker's AdaBoost":"https://pdfs.semanticscholar.org/8d49/e2dedb817f2c3330e74b63c5fc86d2399ce3.pdf" tailored for regression.
-
-In bagging, each training example is equally likely to be
-picked. In boosting, the probability of a particular
-example being in the training set of a particular machine
-depends on the performance of the prior machines on
-that example. The following is a modification of
-Adaboost by Drucker.
-
-Start by selecting a set of training data $n$ and assign to each entry a weight $w_i=1$ for $i=1,2,\dots,n$. As we have done earlier, we could pick say $80\%$ of the data set for training. The algorithm runs as follows:
-o We define the probability that the training sample $i$ is in the set by $p_i = w_i/\sum_iw_i$. We pick $n$ samples (with replacement) to form our training set. We pick a number uniformly in the range $[0,\sum_iw_i]$.
-o We choose then a regression machine (for example plain linear regression or a simple decision tree). A given regression machine makes then a hypothesis.
-o Using every member of the training set with the chosen regression machine we obtain then a prediction $\tilde{y}_i$.
-o We calculate then the loss function $L_i$ for each training sample. We can use various types of loss function as long as we have a value
-$L_i\in [0,1]$.
!split
-===== Gradient boosting: Basics with Steepest Descent =====
+===== Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent =====
Gradient boosting is again a similar technique to Adaptive boosting,
it combines so-called weak classifiers or regressors into a strong
@@ -836,8 +818,8 @@ o Initialize our estimate $f_0(x)$.
o For $m=1:M$, we
o compute the negative gradient vector $\bm{u}_m = -\partial C(\bm{y},\bm{f})/\partial \bm{f}(x)$ at $f(x) = f_{m-1}(x)$;
o fit the so-called base-learner to the negative gradient $h_m(u_m,x)$;
- o update the estimate $f_m(x) = f_{m-1}(x)+\nu h_m(u_m,x)$;
-o The final estimate is then $f_M(x) = \sum_{m=1}^M\nu h_m(u_m,x)$.
+ o update the estimate $f_m(x) = f_{m-1}(x)+h_m(u_m,x)$;
+o The final estimate is then $f_M(x) = \sum_{m=1}^M h_m(u_m,x)$.